UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Implicit and explicit memory: intentional retrieval from different search domains Birt, Angela R. 1997

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_1997-0124.pdf [ 5.02MB ]
Metadata
JSON: 831-1.0087641.json
JSON-LD: 831-1.0087641-ld.json
RDF/XML (Pretty): 831-1.0087641-rdf.xml
RDF/JSON: 831-1.0087641-rdf.json
Turtle: 831-1.0087641-turtle.txt
N-Triples: 831-1.0087641-rdf-ntriples.txt
Original Record: 831-1.0087641-source.json
Full Text
831-1.0087641-fulltext.txt
Citation
831-1.0087641.ris

Full Text

Implicit and Explicit Memory: Intentional Retrieval From Different Search Domains by Angela R. Birt B. A . (Hon.), University of Prince Edward Island, 1994 A thesis submitted in partial fulfillment of the requirements for the degree of Master of Arts in The Faculty of Graduate Studies (Department of Psychology) We accept this thesisas^conforming / to the reo^ri fe^andard/^ The University of British Columbia February, 1997 © Angela R. Birt, 1997 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. The University of British Columbia Vancouver, Canada Department of •ate II rh^^.} tfl7-DE-6 (2/88) Intentions ii Abstract A number of different approaches for considering performance on implicit and explicit memory tests have been proposed. The implicit/explicit retrieval distinction has been conceptualized in terms of (a) multiple, distinct memory systems, (b) differences in the resource demands required by each test (e.g., automatic vs. controlled processing), and (c) differences in the type of processing involved in each test (e.g., data driven vs. conceptually driven processing). All three of these approaches to performance in implicit and explicit memory tests account well for certain aspects of the data and all have difficulty accommodating other aspects. The purpose of this thesis was to explore another approach by which to consider performance on implicit and explicit memory tests. This approach, referred to as the retrieval intentions approach, proceeds from the notion that both implicit and explicit recollection are initiated and guided by intentional, consciously controlled retrieval, but that this retrieval is aimed at different domains of memory. Implicit memory tests target the semantic knowledge domain, whereas explicit memory tests direct search toward the episodic memory domain. In most previous experiments investigating implicit and explicit memory test performance, search domain and familiarity of the search strategy have been confounded. Experiments were conducted in an attempt to tease apart the effects of these two influences on retrieval. In the literature, implicit memory search strategies have been characterized as being familiar and automatic, as opposed to explicit memory search which has been described as effortful. In the present work, the familiarity of memory search strategies was manipulated by using either a familiar or a novel strategy to test whether implicit memory performance can be shifted or influenced by this difference in instructions. The results indicated that performance on an implicit memory task can be influenced by variations in the familiarity of retrieval strategies in the absence of prior study (i.e., baseline) as well as in priming. Explicit cued recall showed a similar pattern of effects. The implications of these results are discussed, along with theoretical prospectives for the retrieval intentions approach to implicit and explicit memory performance. Intentions iii Table of Contents Abstract ii Table of Contents iii List of Tables v List of Figures vi Acknowledgment vii Introduction 1 Theoretical Alternatives 3 Multiple Memory Systems Approach 4 Overview 4 Criteria for separate systems 6 Strengths and weakness of the systems approach 8 Resources Approach 9 • Overview 10 Methods for measuring automaticity 12 Strengths and weaknesses of the resources approach 13 Data Versus Conceptual Processing Approach 16 Overview 16 Strengths and weaknesses of the data versus conceptual processing approach 17 Transfer Appropriate Processing 19 The Retrieval Intentions Approach 21 Overview 21 A difference in search domains 23 Theoretical prospects 24 Testing Intentional Retrieval from Different Search Domains 26 The Present Approach 28 Experiment 1: Semantic Memory R e t r i e v a l 2 8 Intentions iv Method 29 Design 29 Participants : 29 Materials 29 Tests 30 Procedure ; 31 Results and Discussion 32 Experiment 2: Implicit and Explicit Memory Retrieval 44 Method 46 Design 46 Participants 48 Materials 48 Tests 54 Procedure 54 Results and Discussion 56 Word association and stem completion 56 Cued recall. - 63 General Discussion 65 References 68 Appendix A: 79 Appendix B 81 Appendix C: 88 Intentions v List of Tables Breakdown of the Number of Different Responses Produced for Each Cue in the Word Association Task and the Stem Completion Task 34 Actual Number of Indoor Biased, Outdoor Biased, and Neutral Responses as a Function of Test Type and Response Position 43 Layout of Design for Experiment 1 Depicting the Different Experimental Conditions 47 Mean Production Frequency of Each Target from its Cue Under Category Biased Instructions and Under Instructions to Produce Any Words that Come to Mind 49 Production Frequency (%) of Targets from Normative Data Under Baseline and Instructional Category Bias Conditions 52 Priming Measured in Terms of Difference Scores and Relative Scores for the Different Experimental Conditions 60 Intentions vi List of Figures Figure 1: Example of a word association cue with two responses biased for indoors (i.e., sit, bias = 8.3%, table, bias = 11.1%). First responses are shown in the top figure and second and third responses combined are shown in the bottom figure. 36 Figure 2: Example of a stem completion cue with one response biased for indoors (i.e., paper, bias = 12.5%). First responses are shown in the top figure and second and third responses combined are shown in the bottom figure. 37 Figure 3: Example of a word association cue with one response biased for outdoors (i.e., star, bias = 16.7%). First responses are shown in the top figure and second and third responses combined are shown in the bottom figure. 38 Figure 4: Example of a stem completion cue with two responses biased for outdoors (i.e., road, bias = 18.1%, row, bias = 9.8%). first responses are shown in the top figure and second and third responses combined are shown in the bottom figure. 39 Figure 5: Example of a word association cue with no biased responses. First responses are shown in the top figure and second and third responses combined are shown in the bottom figure 40 Figure 6: Example of a stem completion cue with no biased responses. First responses are shown in the top figure and second and third responses combined are shown in the bottom figure. 41 Figure 7: Mean proportion of targets produced under each target bias/test instruction condition on the word association test. (Proportion = number of targets produced / total number of targets), 57 Figure 8: Mean proportion of targets produced under each target bias/test instruction condition on the stem completion test. (Proportion = number of targets produced / total number of targets): 56 Figure 9: Mean number of words recalled for the word cues under the different cue bias/test conditions (i.e., congruent and incongruent) for the explicit cued recall test ' 64 Intentions vii Acknowledgment The author would like to express appreciation to all those who contributed to this study. In particular I would like to thank Amy Siegenthaler and Jennifer Shapka for their help in collecting and analyzing data, and Stephen Porter for his patience, understanding, and useful comments on earlier drafts. Many thanks to my advisor, Peter Graf, for his help, support, and excellent advice. The author of this thesis is supported by a postgraduate fellowship from the Natural Sciences and Engineering Research Council of Canada (NSERC). Intentions 1 Introduction For over a decade there has been growing interest in the distinction between implicit and explicit memory (see Richardson-Klavehn & Bjork, 1988; Roediger, 1990a; Roediger & McDermott, 1993; and Schacter, 1987, for reviews). Implicit and explicit describe the mental sets that guide participants' performance on different memory tests. According to Graf and Schacter (1985, 1987), performance on memory tests such as standard free recall, cued recall, and recognition is referred to as explicit because it requires participants to recollect information from a previous study episode in a conscious, intentional manner. By contrast, performance on implicit memory tests indicates the influence of a prior study episode in the absence of any conscious, intentional attempt at recollecting that episode. Common tests of implicit memory include priming tasks such as word stem completion, fragment completion, and word identification. For example, in a typical word stem completion task participants are presented with a list of targets (e.g., elephant). After an interval, participants are given a set of word stems (e.g., ele ) and instructed to complete each stem with the first word that comes to mind. Implicit memory, or priming, is revealed when participants complete the stems with a greater number of words from the previous target list than with new words to which they had not been exposed in the experiment. A great deal of research has been devoted to comparing the effects of different independent variables on explicit versus implicit measures of memory (see Richardson-Klavehn & Bjork, 1988; Roediger & McDermott, 1993; and Schacter, 1987, for reviews), A number of factors, known to affect explicit memory performance and to have little or no effect on implicit memory performance, have been identified. These include the level of processing of targets at study (e.g., Graf & Mandler, 1984; Jacoby & Dallas, 1981), the extent to which associations are formed between to-be-remembered items (e.g., Graf & Schacter 1989), the age of participants (e.g., Howard, Heisey, & Shaw, 1986; Light & Singh, 1987; Parkin, 1993), and brain damage or disease (e.g., Parkin, 1987; Squire, 1986; Weiskrantz, 1985). There are also a number of variables that affect implicit memory test performance more extensively than.explicit memory test performance. Examples include study-test changes in modality (e.g., Graf, Shimamura, & Squire, 1985; Roediger & Intentions 2 Blaxton, 1987a), study-test changes of various types of surface information (e.g., Kolers, 1975a, 1975b, 1976; Roediger & Blaxton, 1987b), and changes in setting or in the environment (see Graf, 1991). The basic aim of such studies was to establish empirically the distinction between explicit and implicit memory test performance, and a wealth of data exists describing the influence of different variables on these two types of memory retrieval. But what exactly is (are) the defining characteristic(s) that distinguish(es) explicit from implicit retrieval? This question represents the principal challenge for implicit/explicit memory researchers and theorists, and the search for the answer has led to the proposal of several different theoretical approaches One approach ascribes differences between implicit and explicit memory retrieval to the different properties of hypothesized underlying memory systems (e.g., Cohen & Squire, 1980; Squire & Cohen, 1984; Schacter, 1989; Tulving, 1972^ 1983). Another approach posits that dissociations between implicit and explicit memory can be accounted for by differences in resource demands; that is, by automatic versus controlled processing (e.g., Jacoby, 1991, 1993; Jacoby, Ste-Marie, & Toth, 1993). Yet another approach characterizes the implicit/explicit distinction as a distinction between data driven versus conceptually driven processing (e.g., Jacoby, 1983; Roediger & Blaxton, 1987a; Roediger, Weldon, & Challis, 1989). Two of these approaches, the automatic versus controlled processing and the data driven verses conceptually driven processing approaches, have focused on the manner in which information is retrieved. These approaches address the question of how information is retrieved from memory. Another approach, which, surprisingly, has been largely overlooked, is to focus on where memory retrieval is targeted. That is, another possible approach to examining differences between implicit and explicit memory performance is to look at the difference in distinct retrieval domains/categories targeted by each test. Implicit recollection typically requires a conscious, intentional search of the domain of knowledge or language, whereas explicit recollection usually requires a conscious, deliberate search of the domain of a previous event or experience., Recollection within both of these broad domains is always directed at particular sub-domains or categories defined by the specific criteria for retrieval. William James (1890/1962) maintained that we "make search in our memory for a Intentions 3 forgotten idea, just as we rummage our house for a lost object. In both cases, we visit what seems to us the probable neighborhood of that which we miss" (p. 297). In this context a retrieval or search domain refers to the specific "neighborhood" or category of information at which search is directed. It is argued that both implicit and explicit recollection involve a conscious intention or strategy for retrieval, but what differentiates the two is the specific search domain or neighborhood to which retrieval is directed. All of the above approaches have, in some sense, looked at implicit and explicit retrieval in terms of retrieval intentions, for example memory when consciously intending to recollect a specific prior episode versus memory when not consciously intending to remember that prior episode. What exactly is meant by retrieval intentions? The focus of this thesis is on the processing that occurs during memory retrieval—specifically, conscious, intentional retrieval of a specific prior episode1. A retrieval intention refers to the conscious, deliberate goal or plan for initiating and guiding the recollection of information from a previous learning event or episode. It may be that implicit and explicit memory retrieval differ significantly in terms of the different search domains targeted by conscious intentional retrieval in both tasks. Looking at differences in conscious intentions at the time of retrieval in terms of distinct memory search domains provides another avenue for exploring differences between implicit and explicit memory retrieval. The main purposes of this thesis are to provide a general overview of some of the various theoretical approaches that have been proposed to account for implicit and explicit memory test performance, to suggest an additional way in which the implicit/explicit distinction might be approached (i.e., with respect to conscious, intentional retrieval from different search domains), and to report findings from a study investigating this new approach. Theoretical Alternatives Researchers have looked at implicit and explicit memory, test performance in a number of different ways. This section provides a general overview of three distinct 1 Conscious intentions are also an important at the time of study, influencing performance on memory tests. However, encoding intentions were deliberately excluded from discussion in order to keep the focus of analyses specific to intentions during retrieval. Intentions 4 theoretical approaches that have dominated previous work. These approaches will be presented in turn accompanied by a discussion of their strengths and weaknesses in explaining implicit and explicit memory test phenomena. Multiple Memory Systems Approach Overview. According to the multiple memory systems approach, dissociations between performance on explicit and implicit measures of memory are explained by appealing to the existence of different, largely independent, brain systems which are . assumed to operate on different principles and serve different functions. Several different memory system taxonomies have been proposed to account for explicit/implicit memory dissociations and these taxonomic arrangements continue to change with the unfolding of new research findings. For example, Cohen and Squire (1980; see also Squire & Cohen, 1984) distinguished between declarative and procedural memory systems, with the former referring to verbalizable knowledge and the latter to the running off of skilled behavior without the need for conscious recollection. Basically, this declarative/procedural memory system distinction maintains that explicit recollection is a property of, and is supported by, the declarative memory system, whereas, implicit memory phenomena such as skills, habits, and priming depend on the operation of the procedural memory system. The distinction between episodic and semantic memory systems (Tulving, 1972, 1983) has also been employed to explain dissociations between implicit and explicit memory test performance (e.g., Schacter & Tulving, 1982). The episodic memory system is said to be autobiographical, personal, and context sensitive. Episodic memories are assumed to be organized by time and by place of occurrence. By contrast, the semantic memory system is said to involve general, encyclopedic knowledge of the world and language, both of which are not sensitive to temporal or spatial context. In the semantic system information is assumed to be organized hierarchically on the basis of class membership. The episodic memory system is viewed as the basis for explicit remembering, whereas the semantic memory system is considered responsible for performance on implicit memory tasks such as lexical decision, category production, and word completion (e.g., Tulving, Schacter, & Stark, 1982). -. ' Intentions 5 With the recognition that his semantic memory system could not accommodate many of the research findings on priming, especially findings of implicit memory for new associations (e :g., Graf & Schacter, 1985) and of long-lasting priming effects (e.g., Tulving, Schacter, & Stark, 1982), Tulving (1985) later proposed the existence of an additional system, the procedural memory system. Like Cohen and Squire's (1980) procedural memory system, the primary function of this system is the retention of perceptual, motor, and cognitive skills. Information in this system is not declarative in nature, but instead, is demonstrated by doing. Along with the addition of this third memory system, Tulving (1985) proposed a 'monohierarchical' arrangement of the three memory systems, with procedural memory as the most basic system underlying the other two systems (i.e., the semantic and episodic systems). Semantic memory depends on the procedural memory system, and episodic memory is dependent on both. With the proposal of the third memory system, it became less clear which system priming is subserved by. Do priming phenomena depend more on the operation of the semantic system, as Tulving had previously proposed, or do they depend primarily on the operation of the procedural memory system? Tulving (1985) was not entirely clear on the issue of which memory system was responsible for priming phenomena. For example, he suggested that priming effects in word fragment completion might occur in the procedural system, the semantic system, or in yet another system which, because of its ambiguity, he termed Q M for Question Mark. More recently, Tulving's Q M memory system became less ambiguous and was renamed the perceptual representation system (PRS) (e.g., Schacter, 1990, 1992; Tulving & Schacter, 1990). The PRS was proposed in an attempt to account for at least some priming phenomena. Specifically, the PRS is believed to be responsible for priming effects on implicit tests that focus on information concerning the form and structure of words, objects, etc. Representations in the PRS are said to be nonconscious and 'presemantic'. This system is concerned with perceptual information only; it does not process or represent semantic or associative information about stimuli. It is the semantic memory system that is thought to be responsible for priming of more conceptually-based type of information. The PRS was postulated to account only for priming of perceptual-based implicit memory tests, such as stem completion, object decision tasks, and auditory word identification. Intentions 6 Criteria for separate systems. As research developments in implicit and explicit memory phenomena unfold, it seems that the need to postulate additional memory systems to account for the findings increases. Thus, to forestall the unprincipled multiplication of systems, what is required are some criteria for postulating distinct memory systems (see Nadel, 1992; Roediger &Rajaram, 1990; Snodgrass, 1989a). Roediger and Rajaram (1990) outline four criteria that have been used often for distinguishing separate memory systems: (1) independent neural systems, (2) functional dissociation, (3) stochastic independence, and (4) functional incompatibility2. Probably the most important, core criterion for postulating distinct memory systems is the independent neural systems criterion. This criterion maintains that the processing involved in each system must be carried out in different regions or separate circuits in the brain. That is, the systems have different requirements in terms of the information processing strategies they demand, and these different requirements demand different neural circuits. Evidence from patients with selective brain deficits and memory research with animals provide strong support for this criterion. The functional dissociation,or functional independence criterion refers to the. situation in which performance on two different memory tasks varies as a function of an independent variable manipulation (see Dunn & Kirsner, 1988 for a review). Experimental dissociations between performance on two tasks as the result of differential effects of an ; independent variable are often taken as evidence of functional independence of systems. The independent variable can be an experimenter-defined manipulation such as modality of presentation, levels of processing, or context, or a subject variable such as test performance in young versus older adults or amnesic patients versus normal controls. However, the use of dissociation evidence can be problematic when memory systems theories lack set principles that allow one to decide what evidence does and does not address the systems distinction (see McKoon, Ratcliff, & Dell, 1986). 2 These are only four possible criteria for postulating memory systems. Many other criteria have been suggested and utilized in arguing for the existence of distinct memory systems. These four are used as examples because they are representative of some of the more common criteria used to specify separate memory systems. Intentions 7 A third criterion commonly used for postulating separate memory systems is stochastic independence Stochastic independence entails a complete absence of any overlap between operations responsible for differences in.memory task performance. For example, Tulving et al. (1982) demonstrated stochastic independence in a study which had subjects study words, and then examined priming with a word fragment completion test and examined explicit memory with a yes/no recognition test. Stochastic independence was observed in the finding that words successfully completed in word fragment completion were not recognized more successfully than uncompleted words, and vice versa. The functional incompatibility criterion is a fourth criterion often used for postulating separate memory systems. This criterion, which is based on evolutionary principles, requires systems to be specialized to such a degree that the functional problems each system is proposed to handle cannot be handled by another system. Functional incompatibility exists when the function of one memory system cannot, because of its. specialized nature, effectively serve functions of other memory systems. Sherry and Schacter (1987) point out some evidence they believe supports functional incompatibility between systems, such as the distinction between, incremental habit formation, and memory for particular episodes in humans. According to the functional incompatibility criterion, the episodic memory system could not be responsible for the forming of habits and the procedural memory system, could not be responsible for the storing and/or retrieval of episodic memories. This incompatibility exists because each system is assumed to have evolved and adapted a very specialized memory function. The very function that makes a system effective in dealing with one type of information renders it incompatible/ineffective in dealing with other types of information. These four criteria, independent neural systems, functional dissociation, stochastic independence, and functional incompatibility, are only four of many different possible criteria that have been used for postulating distinct memory systems. These criteria are not meant to be exhaustive; they are presented merely as common examples of criteria that often have been used and cited by memory systems theorists when proposing and arguing for the existence of distinct memory systems. More thorough discussions are available elsewhere in the literature (e.g., McKoon, Ratcliff, & Dell, 1986; Nadel, 1992; Roediger & Intentions 8 Rajaram, 1990; Snodgrass, 1989a; Tulving, 1984; Tulving, 1985; Tulving & Schacter, 1990) Strengths and weaknesses of the systems approach. Probably the greatest strength of the memory system approaches is that they do an effective job of accounting for why implicit memory is often preserved in many memory-impaired, brain-injured patients. Evidence for the memory systems explanation has been based primarily on memory dissociations observed in brain-injured patients, most notably amnesic patients. The amnesic syndrome (Hirst, 1982) is characterized by intact perceptual abilities, language, and intelligence together with impaired recall and recognition of recent events and new information. However, despite poor performance on explicit tests of memory, even densely amnesic patients can show entirely normal performance on implicit memory tests (see Shimamura, 1986, 1993, for reviews; Parkin, 1982; Warrington & Weiskrantz, 1968, 1970, 1974). For example, Graf, Squire, and Mandler (1984) found whereas amnesic patients were impaired on free recall, recognition, and cued recall, they showed normal performance on word stem, completion. This type of dissociation observed in amnesic patients has provides some of the strongest empirical grounds for distinguishing between implicit and explicit memory, and for postulating that different memory systems support performance on implicit and explicit memory tasks. A systems account seems like a natural explanation for selective memory deficits in patients with brain damage or disease. It accounts nicely for the fact that one system can be damaged and unable to function, while other systems remain intact and unaffected. Tulving's (1985) 'mo no hierarchical' arrangement of memory systems; which allows for some dependence between the episodic, semantic, and procedural memory systems, provides an account for neuropsychological evidence of independence between memory systems and, at the same time, allows for some neural plasticity when certain systems (or parts of systems) are damaged in the brain. In addition, because the language of the systems approach maps onto the language of the neurosciences, it also provides an important link between cognitive psychology and neuropsychology. This permits easy communication between the different domains of study. Intentions 9 Although, for the most part, the multiple memory systems approach does a good job accounting for performance dissociations between implicit/explicit memory tests in amnesic patients, some of the more recent findings with regard to amnesics prove difficult for it to accommodate. For example, a systems view, especially the declarative/procedural distinction, cannot readily account for reports of amnesic patients showing implicit memory for newly acquired information (e.g., Graf & Schacter, 1985) or for the fact that long-, lasting priming effects (e.g., McAndrews, Glisky, & Schacter, 1987). The fact that amnesic patients can retrieve newly acquired facts and vocabulary even though they do not have any explicit recollection of having acquired the knowledge cannot be attributed to the procedural memory system because the learning of new facts is assumed to be a function of the declarative memory system. Another weakness of the systems approach is that there is no a priori basis for classifying tasks as belonging to one system or another. Thus far, researchers have had to rely primarily on intuition for postulating which memory tasks go with which memory systems. This is most clearly observed in the difficulty that has accompanied deciding which memory system is responsible for priming phenomena; for example, are priming phenomena . subserved by the semantic system, the procedural system, the PRS, or all three? Also, there does not appear to be any real, solid basis for determining a priori which independent variables will affect which system(s) and what kind of effect will result. Other discussions of the application of multiple memory systems view of memory to the implicit/explicit retrieval distinction are available in the literature (e.g., McKoon, Ratcliff, & Dell, 1986; Neely, 1989; Roediger, 1990b; Roediger & McDermott, 1993; Roediger & Rajaram, 1990; Schacter, 1987; Sherry & Schacter, 1987). Resources Approach The multiple memory systems approach represents only one way in which researchers have looked at implicit and explicit memory test performance. Rather than positing different systems, it has been argued that dissociations between implicit and explicit memory test performance might equally well reflect different processes, as in the idea that different processes depend on different quantities of processing resources. Therefore, it is Intentions 10 useful to explore some of the various processing accounts of implicit and explicit memory phenomena. Overview. Another approach to understanding the differences between implicit and ..... explicit recollection assumes that there is a difference in degree of novelty, familiarity, or automaticity of the processing involved in each. It has been argued that information processing can be represented as a continuum in terms of cognitive resource requirements or demands, with unconscious, automatic processing at one end and conscious, controlled processing at the other end (Hasher & Zacks, 1979). It may be that carrying out the demands of one type of memory tasks, such as implicit memory tasks, runs more fluently, more automatically, than carrying out the demands of other types of memory tasks, such as explicit tasks (e.g., Cermak, 1993; Craik & Jennings, 1992). Several lines of research support the idea that implicit memory performance is more familiar and automatic than that of explicit memory, including research on amnesic patients and research on the effects of aging on memory. The first line of research, and probably the best-known, involves experimentation with amnesic patients. As established in the previous section, despite being impaired on explicit tests of memory, amnesic patients can show normal levels of performance on implicit memory tests (see Parkin, 1982; Shimamura, 1986, 1993, for reviews). This finding has been interpreted in terms of the degree of automatic processing involved in implicit versus explicit memory test performance (see Cermak, 1993). It may be that amnesics are able to perform automatic memory processing and this level of processing is sufficient to support their implicit performance. By contrast, strategic, controlled processing may not be available to amnesic patients, and as a result, their explicit memory performance is impaired. Thus, implicit priming for amnesics occurs normally because the processing involved occurs automatically resulting in greater fluency of processing upon repetition (e.g., Cermak, 1993). A second line of research supporting the view that implicit and explicit memory test performance differ in degree of novelty, familiarity, or automaticity comes from studies on memory and aging. It is well known that performance on explicit memory tests declines in late adulthood, with larger declines associated with free recall and cued recall tests than tests of recognifion. In contrast to explicit remembering, aging has little or no detrimental Intentions 11 effects on implicit memory test performance (e.g., Graf, 1990; Parkin, 1993). A number of theoretical views have been proposed to explain normal implicit retrieval in older adults, but one view is particularly relevant to this discussion, the environmental support hypothesis of Craik and colleagues (e.g., Craik, 1983; Craik & Byrd, 1982; Craik & Jennings, 1992). This hypothesis assumes that aging deficits are due to a reduction in available processing resources needed to carry out mental operations. The reduced processing resources of older adults result in a deficit in carrying out self-initiated and effortful retrieval operations such as those involved in explicit memory tests. However, retrieval can be facilitated by maximizing the contributions of external stimulation or environmental support (e.g., cues and instructions provided in a memory test). This is evident in the finding that age-related decrements are typically smaller in recognition than in recall due,to the substantial environmental support involved in recognition memory (i.e., the whole item is re-presented). In contrast to most explicit measures of memory, implicit memory tasks have low self-initiated processing requirements and a high degree of environmental support. For example, in an implicit fragment completion task, retrieval cues are provided (e.g., L E P A N or E L E ) and instructions are given specifying the domain in which to search for responses (e.g., knowledge of language). Implicit test performance is assumed to involve primarily automatic processes, requiring few processing resources. As a result, older adults show little, if any, impairment on implicit memory tasks relative to younger adults. . The conclusions drawn from both the amnesic studies and the aging research rely heavily on the a priori assumption that there has been a reduction in processing resources available to the cognitive system. It is assumed that both the amnesic syndrome and aging result in, or are a consequence of, decreased resources or attentional capacity, and, as a result, controlled processing is severely impaired, but automatic processing remains intact. It is only by accepting these assumptions that the conclusions that implicit memory performance involves automatic processing and explicit retrieval relies on controlled, effortful processing, can follow from these data. Instead of simply making the a priori assumption that performance on implicit tests is automatic, attempts have been made to empirically test whether or not various implicit tests involve automatic processing. Intentions 12 Methods for measuring automaticity. One method for determining whether or not performance on a particular task is automatic is to divide attention between performing that task and some other, potentially distracting, task. This method has been used to test the assumption that implicit recollection is automatic and explicit recollection is effortful and controlled (e.g., Mulligan & Hartman, 1996; Parkin & Russo, 1990; Parkin, Reid, & Russo, 1990; Weldon & Jackson-Barrett, 1993). For example, Parkin et al. (1990) conducted a study which had subjects perform a sentence verification task, with half of the subjects simultaneously engaging in a tone-monitoring task. Memory was tested a day later using either a recognition test or a word fragment completion test on words that were embedded in the sentences. Dividing attention at study had an effect on recognition memory performance, but no effect on primed fragment completion. Another study that employed an attentional manipulation also varied study time for looking at pictures and words (Weldon & Jackson-Barrett, 1993). Dividing attention at study resulted in impaired recall for both words and pictures, eliminated priming for pictures regardless of the duration of exposure at study, and eliminated priming for words presented for a duration of 250 ms, but not when words were presented for 1.5 s or longer Mulligan and Hartman (1996) also examined the effects of divided attention on implicit and explicit memory. They found an impairment for category-cued recall and word-fragment cued recall, and an impairment of priming when it was measured by a category production but not a word fragment completion test. The results of these studies indicate that explicit memory performance is. diminished when attention is divided at study, whereas implicit memory performance can be affected or not affected by divided attention at study depending on variables such as the exposure time of items at study and the type of implicit memory test used (i.e., tests that focus more o f the conceptual or the perceptual properties of stimuli). Jacoby and colleagues (Jacoby, 1991, 1993, 1994, Jacoby, Ste-Marie, & Toth, 1993) have been especially interested in automatic influences on implicit memory test performance, and have developed the process dissociation procedure (PDP) as "a general framework for separating automatic from intentional forms of processing" (Jacoby, 1991, p. 513). The PDP was designed as a tool for estimating quantitatively the separate Intentions 13 contributions of consciously-controlled and unconscious, automatic processes to performance in memory tasks by placing these processes in opposition to each other. In a typical PDP experiment, participants are presented with a set of items (e.g., word list) under different study conditions, and tested under either of two instructional conditions: inclusion and exclusion. In the inclusion condition, participants are instructed to respond to cues on the test with items from the list or, failing that, with any item that comes to mind. Responding in this condition involves either intentional inclusion of words from the study list or unintentional, automatic production of words in response to the cues. In the exclusion condition, participants are told to respond to the test cues with only items that were not in the previous study list. The different instructions produce a situation in which automatic retrieval processes promote responding with studied items in the exclusion condition, whereas intentional retrieval processes inhibit responding. Therefore, any previously studied items produced in the exclusion condition are presumed to provide evidence for automatic processing. A n increase in errors produced by such a manipulation is taken as evidence that intentional processing was reduced and automatic processes were left unopposed. Similar to other theorists who equate implicit memory performance with automatic processing (e.g., Cermak, 1993; Craik & Jennings, 1992), proponents of the PDP suggest that this tool demonstrates that processing on implicit memory tests is primarily automatic. Jacoby (1991) maintains that the implicit versus explicit retrieval distinction is analogous to. the automatic versus controlled processing distinction in the attention literature. Automatic processing is characterized as being fast, passive^ not necessarily accompanied by conscious awareness, unintentional, and not requiring processing capacity (e.g., Hasher & Zacks, 1979; Posner & Snyder, 1975). Jacoby (1991) argues that implicit memory test performance is essentially automatic because characteristics similar to automatic processing • have been attributed to implicit memory (but see Komatsu & Graf, 1994; Graf, Komatsu, & Uttl, 1995). Strengths and weaknesses of the resources approach. The theory that implicit test performance involves primarily automatic processing, whereas explicit test performance involves more effortful, controlled processing does a good job accounting for the selective : Intentions 14 memory deficits observed in brain injured/diseased patients, for the general decrease in explicit, but not implicit, test performance accompanied by aging, and for the fact that the capacity for implicit remembering develops, and is fully functional^ at an earlier age than the capacity for explicit remembering (e.g., Greenbaum & Graf, 1989). If implicit retrieval is mediated by processes that are unconscious, noninteritional, and require few, if any, processing resources, and explicit retrieval is mediated by conscious, intentional, resource-demanding processes, then it makes sense that implicit retrieval develops very early in life and is immune to age changes and brain damage/disease, whereas explicit retrieval shows the opposite effects. As with the multiple memory systems approach, the differences in resource demands approach has difficulty accommodating all of the implicit/explicit research findings. Especially problematic is that not all aspects of implicit memory test performance appears to be automatic. For example, performance on implicit tasks such as word fragment completion (e.g., Weldon, 1993) and solving word anagrams seems to require considerable effort, especially when the cues have few or only a single solution (e.g., ss ss_ for assassin, or ni n for onion). The phenomenon of implicit memory for new associations (e.g., Graf & Schacter, 1985, 1987; Schacter & Graf, 1986) also suggests that implicit memory performance is not always automatic. This phenomenon shows that the normally 'activated' response to a cue can be altered and controlled by a single previous encounter. For instance, a single encounter of two unrelated words (eg:, window: shirt) during study is enough to increase the likelihood that the word 'shirt' will be produced in response to the test cue 'window', even though the word 'pane' (for example) and the cue word 'window' are typically more closely related semantically than 'shirt', and 'window'. It does not seem likely that a single presentation of an unrelated pair of words would be enough exposure to render the recollection of the association between the two words automatic. Therefore, retrieval in some implicit tests does not appear to be determined by the degree of practice/automaticity or associative strength between the cue and target, but by conscious, intentional processing specified by the instructions and/or cues instead (see Lockhart, 1989). It also has been found that both implicit and explicit memory for newly acquired associations depends on elaborative study processing (e.g., Graf & Schacter, 1985; Schacter Intentions 15 & McGlynn, 1989) and that subjects can adopt different encoding strategies to enhance implicit memory performance (e.g., Neill, Beck, Bottalico, & Molloy, 1990). In these cases, implicit memory retrieval resembles controlled, explicit retrieval more than retrieval guided by automatic processes. These research findings suggest that not all performance on implicit memory tests is automatic and that conscious, intentional retrieval plays a significant role in what is remembered on both implicit and explicit tests. Probably the easiest way to establish whether or not conscious, effortful, intentional retrieval is actually present in implicit memory performance is to simply ask a participant taking part in an implicit memory task; ask him or her whether or not they engaged in any consciously controlled, deliberate, intentional processing in completing the task. Take category production as an example. If participants are asked to name the first six mammals that come to mind, they will most assuredly admit to a deliberate, intentional search of their knowledge of mammals (i.e., semantic memory). Similarly, in word stem completion, word fragment completion, and in solving anagrams participants must intentionally search their knowledge of language in order to generate words in response to word stems (e.g., K A ), word fragments (e.g., _ H _ _ T _ H ) , and scrambled words (e.g., Y A R M = ) respectively. The important question concerning the automatic versus controlled processing debate may not necessarily be whether or not performance on all implicit is automatic, but a question of which aspect of the implicit test is executed by automatic processing. It may be that implicit memory test performance is automatic only in the sense that they do not involve controlled recollection from a specific prior episode. It may be the retrieval of information from a previous learning episode that occurs relatively automatically. However, implicit test retrieval is consciously controlled in the sense that the processing that initiates and guides performance on the memory test is not automatic; it requires intentional, strategic retrieval. Like the multiple memory systems approach to explaining implicit and explicit memory phenomena, the resources approach accommodates certain aspects of the data well, but has difficulty accounting for all of the research findings. Intentions 16 Data Versus Conceptual Processing Approach Overview. Another approach for explaining implicit and explicit memory phenomena maintains that the difference between implicit and explicit retrieval is really a difference between data driven and conceptually driven processing (e.g., Jacoby, 1983; Roediger & Blaxton, 1987a, 1987b). Data driven, or bottom-up, processing refers to processing that is initiated and guided by physical, perceptual, surface features of stimuli in a memory task, whereas conceptually driven, or top-down, processing refers to processing that is initiated and guided by encoded meaning of concepts, or to semantic processing, elaborative encoding, organization, mental imagery, etc. The chief propbnents o f this view are Roediger and colleagues (e.g., Roediger & Blaxton, 1987a, 1987b; Roediger & Srinivas,T993; Roediger, Weldon, & Challis, 1989; Srinivas & Roediger, 1990). Although they assume that most tasks require a mixture of both types of processing, they have argued that memory tasks can be differentiated according to whether they rely more heavily on data driven or conceptually driven processing. According to this view, most standard implicit tests tap primarily data driven processing, whereas explicit memory tests tap processing that is predominantly conceptually driven. As a result, conceptual processing during memory retrieval will have little effect on implicit test performance, and perceptual processing during retrieval will not affect performance on explicit tests. The data driven/conceptually driven processing distinction was formulated out of a need to account for contradictory research findings concerning variables that usually produced dissociations between implicit and explicit retrieval. Manipulations that ordinarily dissociated between implicit and explicit memory performance failed to show the regular pattern of results with some implicit tests of memory. For example, generating words during study generally produces better performance on explicit tests than simply reading words, but the opposite occurs on implicit tests (i.e., reading words is better than generating words). However, Blaxton (1989) found a generation effect for a general knowledge implicit test (e.g., subjects answer general knowledge questions, the answers to some of which are presented in the study phase; for example, "What did Socrates drink at his execution?" Answer: Hemlock), the same pattern of results as found for explicit tests Similarly, it has been found that a levels of processing (LOP) manipulation affects explicit, Intentions 17 but not implicit, memory test performance. Yet, some studies (e.g., Srinivas & Roediger, 1990) have found that implicit tests such as category production (e.g., subjects are asked to produce exemplars to given category names, with no reference made to the study list which included some of the exemplars) show results that more closely resemble explicit test performance. Accordingly, unlike the usual finding that study-test modality shifts produce a strong effect on implicit, but not explicit, test performance, Blaxton (1989) found that such shifts have no effect on priming of general knowledge questions (see also Srinivas & Roediger, 1990). Because memory tests such as general knowledge questions and category production (e.g., Hamann, 1990) more closely resemble explicit tests than implicit tests in their performance effects, they have been called implicit conceptual tests. This means that not only can dissociations be demonstrated between implicit (primarily perceptual) and explicit memory test performance, but they can also be found within implicit memory tests; that is; between perceptual and conceptual implicit test performance Likewise, dissociations can be demonstrated within explicit memory tests. Examples of possible perceptual explicit tests include having to remember the modality or. the type font in which a previous study item was presented or remembering only the items from the first half of the study list. It could be argued that the existence of conceptual implicit tests and perceptual explicit tests provides evidence that the data driven versus conceptually driven distinction does not capture fully the difference between the implicit versus explicit test distinction, and therefore, does not provide an adequate explanation for dissociations between implicit and explicit retrieval. However, Roediger and associates (e.g., Roediger & McDermott, 1993; Roediger, Weldon, & Challis, 1989) have stressed that the data driven/conceptually driven contrast is not meant to be entirely co-extensive with the implicit/explicit distinction. Strengths and weaknesses of the data versus conceptual processing approach. This approach accounts well for the bulk of performance dissociations in implicit and explicit memory tests. It provides a useful classificatory scheme in which memory tests can be organized, and can predict which tests will or will not be influenced by the manipulation of different variables. For example, conceptual implicit and conceptual explicit memory test performance should be influenced by levels of processing and generate versus read study Intentions 18 manipulations, whereas perceptual implicit and perceptual explicit memory test performance should not show any effects due to these factors. Conversely, perceptual implicit and perceptual explicit, but not conceptual implicit and conceptual explicit, recollection should be affected by changes in modality and changes in surface features of items between study and test. By distinguishing between data driven and conceptually driven implicit memory . tests, this approach is also good at accommodating observations such as persistent priming effects on certain implicit memory tests, implicit memory for newly acquired associations, and the effects of context on some implicit tests. Most of these observations were made when conceptually driven implicit memory tests were used. These findings map well onto similar findings using traditional (i.e., conceptual) explicit tests of memory because conceptual implicit test performance and explicit test performance tend to be influenced by the manipulation of different variables in very similar ways, (see Schacter, 1987). One problem for the idea that perceptual and conceptual processes account for dissociations between implicit and explicit memory, test performance comes from evidence of perceptual variables affecting conceptual tests (e.g., Hunt & Toth, 1990) and conceptual variables affecting perceptual tests (e.g., Bassili, Smith, & MacLeod, 1989; Hunt & Toth, 1990). To illustrate, Hunt and Toth (1990) found that orthographic distinctiveness (a perceptual variable) influenced performance on both a perceptual implicit test and a conceptual explicit test. Craik, Moscovitch, and M c D o w d (1994) found explicit word-stem cued recall to be equally sensitive to a modality switch between presentation and test as. implicit word fragment completion and implicit word stem completion. Performance on word stem and word fragment completion tasks (data driven tasks) has been found to be facilitated by semantic as opposed to nonsemantic study tasks (e.g., Challis & Brodbeck, 1992; Squire, Shimamura, & Graf, 1987), and they also have been found to be facilitated by cross-modal presentation to some extent (e.g., Graf & Schacter, 1985; Roediger & Blaxton, 1987a). In addition, although recognition memory is considered to be a highly conceptually driven task, it has been well documented that recognition test performance is sensitive to changes in surface features of the stimuli (eg:, Kirsner & Smith, 1974; Jacoby & Dallas, 1981). If it is the case that (a) memory test performance is dependent on the degree to Intentions . 19 which the type of processing (conceptually driven versus data driven) at study is recapitulated at test, and (b) variations in conceptual processing have little effect on most implicit tests and variations in surface features have little effect on conceptually driven tests (e.g., Roediger et al., 1989), then findings like those listed above are problematic. , Another weakness of this approach is that it has trouble accounting for certain phenomena in amnesic patients. As Roediger et al. (1989) point put, the most natural argument for the general findings in amnesic patients would be that they show preserved priming on data-driven tests of memory, but not on conceptually driven tests. That is, their memory impairment is largely a deficit in conceptual, elaborative processing. This data-driven hypothesis of preserved priming seems to work well for most tasks in which amnesics show, intact priming, such as word fragment completion, picture completion, perceptual identification, and lexical decision. However, the finding that amnesic patients show intact priming on implicit memory tests that seem to be conceptually driven, such as free associating to category names and word completion for new associates (see Shimamura, 1986, for a review), poses a problem for this approach. Therefore, this approach shed light on the full range of tasks that demonstrate priming in amnesic patients. Also, if most explicit tests are conceptually driven, it follows that no dissociations should be found in neither amnesics nor normal people between conceptually driven implicit and explicit tests. To summarize, the conceptually driven versus data driven processing distinction serves as a useful classificatory scheme for categorizing a wide range of memory tests. It does a great job organizing implicit and explicit memory tests according to the type of processing involved in each, and as a result can aid in predicting how different variables might affect performance on the implicit and explicit tests, depending on whether retrieval on the test is classified as primarily data driven or conceptually driven. However, similar to the previous two approaches discussed, this approach cannot account for all of the research findings, especially those involving amnesic patients. Transfer Appropriate Processing Intentions . 20 It is important to note that the processing approaches outlined above (the automatic versus controlled processing approach and the data driven versus conceptually driven processing approach) are particular instantiations.of a more general principle of memory. Both processing approaches build on a broader assumption about memory known as Transfer Appropriate Processing (Morris, Bransford, & Franks, 1977) or, similarly, Transfer Appropriate Procedures (TAP) (Roediger & Blaxton, 1987a, 1987b; Roediger & Weldon, 1987; Roediger, Weldon, & Challis, 1989). This framework, rooted in Kolers' procedural viewpoint (Kolers, 1975a, 1975b, 1979; Kolers & Roediger, 1984), Tulving's encoding specificity hypothesis (Tulving & Thomson, 1973), and the work of Morris and colleagues (Bransford, Franks, Morris, & Stein, 1979; Morris et al., 1977), does not focus as much on the memory tasks themselves as it does on the match in mental operations performed at study and test. According to the logic of TAP, memory test performance will be facilitated to the degree that the same, or similar, type of mental operations are used as those engaged during previous study. That is, memory performance is determined by the extent of overlap between study and test processing. Numerous memory studies (see Roediger & Srinivas, 1993) have demonstrated the TAP principle, including implicit memory test studies which show that priming effects are larger when the same as opposed to different processing (e.g., data driven versus conceptually driven processing; automatic versus controlled) is used at study and at test. For example, priming is generally larger for words that are studied and tested in the same versus a different sensory modality (e.g., visual-visual versus auditory-visual) (e.g., Clarke & Morton, 1983; Roediger & Blaxton, 1987a; Schacter & Graf, 1989), when study and test items are presented in the same rather than different symbolic forms (as in the case of words versus pictures) (e.g., Weldon & Roediger, 1987), or in the same versus different languages (e.g., Durgunoglu & Roediger, 1987a; Watkins & Peynircioglu, 1983). Although TAP serves as a good guiding principle for memory research, it does not, by itself, provide an explanation for the similarities and differences between implicit and explicit memory effects. It is a mistake to use TAP as an explanation of performance dissociations between memory tests. TAP is a useful framework for thinking about memory test performance and dissociations. It is an assertion about human memory that serves as an Intentions 21 organizational principle for theorizing about specific performance dissociations, but does not make any special claims about explaining their existence. However, by following the logic of this framework and using it as a starting point in investigating memory phenomena, it may be possible to come closer to a meaningful explanation of implicit/explicit memory effects. The Retrieval Intentions Approach Thus far we have looked at three of the ways in which performance dissociations between implicit and explicit memory tests have been approached. This review is not meant to be exhaustive; its purpose is to give a flavor for some of the ways in which implicit and explicit research findings have been considered. There are many ways in which the implicit/explicit distinction has been conceptualized, and these approaches represent only three Are there other important ways of looking at implicit versus explicit retrieval that have not yet been investigated? Much of the previous work exploring the differences between implicit and explicit recollection has been concerned with the manner in which information is retrieved. A n alternative approach is to concentrate on differences in where memory retrieval is targeted. That is, instead of focusing on how implicit and explicit memory retrieval is carried out (e.g., automatic versus controlled retrieval), it is possible to focus on the particular search domains that are targeted by implicit and explicit retrieval. Implicit recollection typically requires a conscious, intentional search of the domain of knowledge or language, whereas explicit recollection usually requires a conscious, deliberate search of the domain of a previous event or experience. It is argued that both of these types of recollection involve a conscious intention or strategy for retrieval, but what differentiates the two is the degree of familiarity or automaticity of the retrieval strategy and the specific search domain to which retrieval is directed. Overview. Implicit and explicit memory test performance were defined in terms of the conscious, intentional processing that initiates and guides retrieval in both tests (Graf & Schacter, 1985, 1987). As specified previously, memory tests such as recall and recognition are referred to as explicit because participants try to recollect information from a specific prior study episode in a conscious and intentional way. By contrast, priming tests such as Intentions 22 word stem completion and fragment completion are called implicit because there is an influence of a previous specific episode on performance in the absence of a conscious, intentional attempt to recollect that episode. These definitions can be conceptualized in two different ways (see Richardson-Klavehn & Bjork, 1988; Schacter, 1989). First, implicit and explicit memory test performance could be considered in terms of the phenomenological experience (i.e., conscious versus unconscious) that follows intentional retrieval on each memory task. That is, what may differentiate explicit from implicit retrieval is the presence or absence of an "awareness of remembering" from a previous event or episode, which is typically a consequence of performance in explicit, but not implicit, memory tests. The investigation of the phenomenological experience resulting from implicit and explicit memory retrieval is concerned with whether memory test subjects are aware or not aware of retrieval from a prior study episode. The aware versus unaware retrieval distinction is largely a measurement issue. It is focused primarily on trying to develop ways of measuring whether or implicit memory tests are contaminated by the awareness that memory retrieval is from a prior study episode. A second interpretation of the definitions of implicit and explicit memory retrieval could focus on the intentional versus nonintentiohal retrieval distinction, and as a result, the conscious awareness that accompanies the initiating and guiding of retrieval would be interpreted as the critical difference. In other words, implicit and explicit memory test performance could be considered in terms of the conscious intentions that direct memory retrieval. In both tests, performance depends on the conscious intentions for retrieval, but they differ in whether there is an intention to retrieve information from the domain of general knowledge or an intention to retrieve information from a previous event or experience. To summarize, Graf and Schacter's (1985, 1987) definitions of implicit and explicit remembering can be interpreted either in terms of subjects' awareness (or lack thereof) that retrieval is from a prior study episode or in terms o f the differences in retrieval intentions for searching different memory domains. Schacter, Bowers, and Booker (1989) argue that it is preferable to distinguish explicit from implicit memory in terms of intentional versus Intentions 23 unintentional recollection of a previous episode rather than in terms of the presence or absence of conscious "recollective. experience" resulting from retrieval from that episode, primarily because rigorous criteria (e.g., the retrieval intentionality criterion) can be developed for making the former, but not the latter, distinction. Ascertaining whether subjects do or do not exhibit any conscious awareness of remembering items from a previous study episode during an implicit test is extremely difficult to do, other than asking them after they finish the test. A difference in search domains. Many views of memory maintain that memory retrieval is a conscious, intentional process that selectively targets particular domains of knowledge (e.g., Humphreys, Bain, & Pike, 1989; Nelson, 1989; Raaijmaker & Shiffrin, 1981; Seidenberg & McClelland, 1989; Shiffrin, 1970). The appropriate search domain for a memory task is typically specified by instructions, test cues (e.g. word stems), contextual cues, and response requirements. As described earlier, for explicit memory tests, participants are directed to search particular recent episodes or experiences (i.e., episodic memory domain), whereas for implicit memory tests, participants are asked to perform a task that requires a search for, and a demonstration of, knowledge or skills (i.e., semantic or procedural memory domain). Not only can implicit and explicit recollection involve a conscious, intentional search of different memory domains, but specific to-be-searched memory domains or 'neighborhoods' can be specified within each. For example, a search of the episodic memory domain can be further specified by instructions to not only recollect words from a particular study list, but to recollect only the first few words on the list or only those words written in red ink, for example. Similarly, a search of the semantic memory can involve, for example, the retrieval domain or category of the first response that comes to mind or it can be more focused, as in the retrieval of only words that begin with the letters H Y . Could it also be the case that subjects use different strategies for searching the different domains targeted for retrieval? A search of the semantic memory domain is often a search for familiar/fluent information such as language or general, everyday knowledge. By contrast, a search of the episodic memory domain is typically a search for less familiar, very specific information with a particular spatial/temporal context. For example, the Intentions 24 explicit recollection of particular items from a previously presented list in a cued recall test is not familiar or automatic. It requires a conscious, deliberate "thinking back" and search of the study list. On the other hand, implicit recollection of items from a previous study list in a free association task appears to be more familiar and fluent. It requires a conscious search of semantic memory in order to retrieve items that are associated with the cue words on the test. However, there is nothing in the definition of a free association test that specifies it must be that way. It may be that manipulating the criteria for retrieval in implicit tests (i.e., specify a general search domain versus a narrow search domain) can increase or decrease the degree of fluency or familiarity of recollection in such tests. In most previous experiments investigating implicit and explicit memory test performance there has been a confound between search domain and search strategy familiarity. Explicit memory tests require a search of a unique, episodic domain and the search strategy is typically effortful because subjects have not had practice retrieving from that domain. By contrast, implicit tests require a search of the familiar domain of general knowledge. The strategy for searching this domain is familiar and well-practiced. The retrieval intentions approach proposes that by looking at implicit and explicit test performance in terms of search domain and search strategy familiarity, and teasing apart the effects of each on memory performance, it is possible to gain further insight into performance on implicit and explicit memory tests. Theoretical prospects. A potential strength of the retrieval intentions approach to implicit and explicit memory performance is that it provides a new, largely unexplored, way of looking at memory phenomena. Many of the research findings concerning implicit and explicit recollection can be reconsidered in terms of consciously controlled retrieval from different search domains. For instance, the influence of various experimental variables, the early development of the capacity for implicit memory, and the differences in memory i performance between young versus older adults and amnesic patients versus normal controls. To illustrate, the finding that amnesic patients can show normal priming effects, while, at the same time, show severely impaired performance on explicit tests of memory can be interpreted according to this view that implicit and explicit retrieval is aimed at Intentions 25 different memory domains. Instead of accounting for the selective memory deficit observed in amnesic patients in terms of differential.impairment in memory systems or in terms of impaired controlled, but intact automatic, processing, it may be that amnesic patients suffer from an inability to initiate, guide and control retrieval to a specific search domain. It is possible that, regardless of the particular search domain targeted, amnesic patients do not have a difficulty in modifying and adapting information they already have in memory, but do have difficulty in consciously, intentionally retrieving new information. If this were the case, one would expect amnesic patients to show a deficit in consciously controlled recollection from both the episodic and the semantic memory domains. To summarize, both implicit and explicit memory test performance involve the operation of conscious intentions. Definitions of implicit and explicit recollection (Graf & Schacter, 1985, 1987) use the term 'intention' with reference to whether or not participants consciously intend to retrieve information from a specific prior episode. However, intentional processing, in a more general sense, plays a definite role in governing memory test performance. Retrieval intentions, whether they be subject initiated or experimenter provided, serve to initiate and guide memory test performance. When all aspects of the memory test situation are held constant with the exception of instructions, one way in which retrieval intentions can differ between implicit and explicit tests is that they are aimed at different search domains, with explicit memory retrieval involving a search of the episodic memory domain and implicit memory retrieval being directed at primarily the semantic memory domain when a search of general knowledge or language is required or the procedural domain when demonstrating habits or skills. The automatic versus controlled processing distinction does not seem to capture fully the differences between implicit and explicit recollection because it is apparent that not all aspects of implicit memory test performance is automatic, but is often guided by consciously controlled retrieval mechanisms. However, unless the effects of the degree of familiarity/fluency/ automaticity of retrieval strategies is disentangled from the effects of intentional retrieval from different search domains, it will be difficult to determine which is the key factor in contributing to differences in implicit and explicit memory performance. Intentions 26 One approach to separating memory search domains from fluency/familiarity of processing is to manipulate the criteria for retrieval. If specifying more constrained criteria (e.g., a narrower search domain) for implicit memory retrieval influences performance, than it cannot be the degree of fluency, but the different search domains targeted, that is critical in differentiating implicit from explicit memory test performance. Testing Intentional Retrieval from Different Search Domains How can it be determined whether or not conscious, intentional retrieval from different search domains can provide any account for implicit and explicit memory test performance? Most experiments demonstrating performance differences in normal subjects between implicit and explicit memory tests have manipulated variables associated with either the test cues, the instructions, the context, or the type of response required. All of these test properties influence the subjects' intentions/strategies for retrieval at the time of test. As discussed previously, the search domain that is targeted in implicit versus explicit tests may be an important property of retrieval intentions/strategies, however search domain is almost always confounded with the degree of familiarity of the search strategy. By varying instructions (and holding context and required response constant) at the time of test to manipulate the fluency/familiarity of the strategy for retrieval from a specific search domain, it may be possible to learn more about the influence of different search domains on implicit and explicit retrieval. Graf and Birt (1996) recently reported a preliminary experiment that explored the ; effect of different retrieval criteria on implicit and explicit memory test performance. Specifically, a pilot study was conducted that focused on the effect of retrieval from different search domains targeted by implicit and explicit memory tests. The purpose was to determine the influence on implicit and explicit test performance due to different to-be-searched memory domains, and whether or not specifying different criteria for retrieval from these domains affects performance on these tests. If it were found that performance on the implicit test can be shifted around by different retrieval strategies requiring more or less degrees of effort, then this would show that implicit recollection is not always familiar/automatic. Intentions 27 Subjects were shown a list of 24 words, with two words belonging to each of 12 different categories (e.g., a human body part, a musical instrument, etc.). Al l words were atypical instances of their respective categories (e.g., throat for a body part, and bassoon for a musical instrument). A levels of processing (LOP) manipulation was employed at study; some items were studied semantically and other items were studied nonsemantically. At test, subjects received either an implicit category production test or an explicit cued recall test, with the same cues (category labels) for both tests. For the implicit test, subjects were asked to. produce the first six exemplars that came to mind in response to the category label, whereas for the explicit test subjects were asked to try to remember instances from the previous study list that fit under the particular category label. The most important manipulation—the search domain manipulation—was implemented via instructions that directed subjects to generate/recall "any" instances belonging to each category (large search domain), or to recall/produce only "typical" exemplars for each of the categories (smaller, more narrow search domain). Because the instructions to retrieve only typical instances constrain the search domain more than the instructions to anything retrieve any instances, it was hypothesized that performance would be higher in the anything condition than in the typical retrieval condition for both implicit and explicit tests of memory. Although the results of thesis were preliminary, because baseline performance was quite low and there were too few subject per condition, they were encouraging. Consistent with previous findings, the L O P study-task manipulation had a greater effect (in terms of effect sizes) on cued recall than priming. More importantly, the results pointed to the possibility of an interaction (although not statistically significant) between search domain and the levels manipulation. For cued recall, the anything condition showed higher recall than the typical condition, but only for those words that were studied semantically. For the category production test, priming was slightly higher in the anything than the typical condition, but this occurred only for words studied nonsemantically. These findings were promising because they suggested that implicit memory performance can be decreased by specifying a limited/constrained search domain for retrieval. Although these results were preliminary, they were interesting enough to warrant further research. At the very least, they give some indication that the difference in search domains, and the degree of familiarity Intentions 28 of the strategy for retrieving information from those domains, may contribute to performance on implicit and explicit tests of memory. The Present Approach To follow up on the suggested findings from this exploratory experiment a new but similar method was developed to try to de-confound the effects of intentional retrieval from different search domains and the familiarity of the search strategy on performance in implicit versus explicit memory tests. The familiarity of retrieval strategies/intentions can be manipulated readily via test instructions. By specifying distinctive subject categories (e.g., indoors versus outdoors) from which responses to each cue are to be generated, it is possible to place additional constraints on subjects' intentional search for a response thereby making it more effortful, as opposed to when no additional retrieval constraints are imposed (e.g., any category). Experiment 1: Semantic Memory Retrieval The purpose of Experiment 1 was to determine whether performance on a semantic memory task is affected by different retrieval strategies as specified by task instructions. Before the influence of consciously controlled retrieval from different knowledge domains on implicit and explicit memory test performance was tested, it was. necessary to demonstrate first that the familiarity of semantic memory search can be affected by instructions to retrieve information from different domains (i.e., different subject categories), in the absence of any prior study episode. This first experiment examined semantic memory retrieval using two different tasks: word association and word stem completion. In a typical word association task, participants are presented with a cue word (e.g., C O L D : ) and are asked to respond with the first word(s) to come to mind. On word stem completion tasks, participants are presented with word stems (e.g., ST _, APP ) and are required to complete the stems with the first words that come to mind. Intentions 29 These particular tasks were selected for a number of reasons. For instance, they are advantageous over other types of implicit memory tests, such as word identification, picture completion and lexical decision tests, because they are much easier and more economical to administer (i.e., paper and pencil tests can be used), they are readily applied to different patient populations (e.g., young children, older adults, Korsakoff patients, etc.), and, importantly, they are very similar to tests used for assessing explicit memory, such as free and cued recall. Word association and stem completion tasks need to differ from cued recall only in terms of instructions (see Graf & Mandler, 1984; Schacter & Graf, 1986). The minimal difference between these types of memory tests makes differences between implicit and explicit retrieval easier to interpret. Method Design. The experiment was a 2 x 2 factorial with both task type (word association, stem completion) and instruction type (anything, indoors, outdoors) as within-subjects factors. The two tests were counterbalanced with the different instruction conditions across participants. Participants. Data were obtained from a sample of 222 student volunteers from the University of British Columbia who participated in the project in return either for pay ($5) or for credit in an introductory psychology course. However, because several subjects demonstrated fluency problems with the English language, only 216 were included in the final data analyses. O f these 216 participants, 71.3% were female and 28.7% were male, the mean age was 20.18 years and they averaged two years of university education. Materials. A total of 150 different cues were presented in the experiment: 75 word cues (e.g., Chair ) for the word association task and 75 two-letter word stems, (e.g., R O ) for the stem completion task. The cues were grouped into three lists of 25 cues per list for each of the tasks—word association and stem completion (i.e., 3 lists x 25 items x 2 tasks = 150 items). The word cues and the word stem cues were selected using different criteria. The word cues, used in the word association task, were taken from Palermo and Jenkins' (1964) word association norms. Only those cues that had four or more different responses with Intentions 30 production frequencies of 15 or greater out of 500 were included. The mean length of the cue words was 5.15 letters (range = 3 to 9). The two-letter word stems, used in the stem completion task, were selected on the basis that for each word stem a pocket English dictionary had to list at least 10 common words with the same stem (e.g., food, forest, fountain, foot, fox, foliage, fork, fog, etc.), thus ensuring that each subject would be able to easily generate completions.for each stem. The North American Adult Reading Test ( N A A R T ) (Nelson, 1982; Nelson & O'Connell, 1978; Spreen & Strauss, 1991) was used as a test of fluency in the English language, following the administration instructions of Spreen and Strauss (1991). The N A A R T requires subjects to pronounce 61 different irregular, rare English words, and it was used as a means of identifying and screening out participants who did not meet the criterion for fluency in the English language (i.e., estimated verbal IQ less than 95). Tests. Data were collected using two standard implicit tasks: word association and stem completion. All cues were displayed individually and in succession on a computer screen. Al l cues stayed on the screen until a response was typed and the return key was pressed. For the word association task, participants were presented with a word cue (e.g., Mountain: _ _ _ _ _ _ _ ) , whereas for the stem completion, they were presented with two-letter word stems as cues (e.g., ST ). For all cues, subjects were asked to generate three different responses that: (a) fit the cues provided, and (b) conformed to the test instructions specifying different categories for retrieval (i.e., indoors, outdoors, anything). The two subject categories under which participants were asked to generate responses were 'indoors' and 'outdoors'. That is, participants were instructed to respond to the cues with words associated with an object or activity typically occurring either indoors or outdoors. The subject categories 'indoors' and 'outdoors' were selected because they represent broad, largely independent, but somewhat overlapping categories. Also, these categories are very familiar, which is important to ensure that most participants would have little difficulty producing responses for the cues according to these general subject categories. Subjects also were asked to generate responses under an 'anything' condition in which they were instructed to respond to the cues with the first words that came to mind. Intentions 31 Procedure. The word association and stem completion tasks were administered separately in blocks, one after the other. Within each task, the cues were subdivided into three different lists. Within each of the three lists the cues were presented in random order to each participant. Therefore, participants were presented with one third of the cues under each instruction condition (anything, indoors, outdoors) for both tasks (word association and stem completion). The instructions and list order for the cues were counterbalanced across participants, with the cues presented in blocks. That is, all participants completed all the word cues before moving on to the word stems, or vice versa. Instructions were given prior to each task (i.e., instructions for completing the word association vs. the stem completion task) and within the tasks before each block of cues (i.e., instructions specifying the subject category for retrieval: indoors vs. outdoors vs. anything). For each participant, the data were collected in a single session that lasted an average of 50 min. Participants were tested individually with the experimenter present. Three practice trials were provided for each instructional/cue list condition. For each of the 150 trials (75 trials per task type), one cue — either a word cue or a word stem cue ~ was displayed on a computer monitor three times in succession, and the participants' task was to produce a distinct response each time it was presented, according to the instructions for each condition. In the anything condition, the instructions were to say simply the first three words that came to mind in response to each cue. Specifically, subjects were told: "You are going to be presented with a series of words (or word stems) one at a time. Each cue will be presented three times in a row, and your task is to say the first word that come to mind each time the cue is presented. The three responses you give to each cue should be distinct. Please respond as quickly as possible and say the first word that pops into your mind." In the indoors condition, the instructions were to respond to each cue with'the first three words associated in some way with indoors that came to mind. The exact instructions were: "You are going to be presented with a series of words (or word stems) one at a time. Each cue will be presented three times in a row, and your task is to say the first word related to indoors that come to mind each time the cue is presented. The three responses you give to each cue should be distinct. Please respond as quickly as possible and keep in Intentions 32 mind that your responses should be associated in some way with indoors." By contrast, in the outdoors condition, the instructions were to respond to the cue with the first three words that were related in some way to the outdoors. The specific instructions given for this condition were: "You are going to be presented with a series of words (or word stems) one at a time. Each cue will be presented three times in a row, and your task is to say the first word related to outdoors that comes to mind each time the cue is presented. The three responses you give to each cue should be distinct. Please respond as quickly as possible and keep in mind that your responses should be associated in some way with outdoors." Each participant had approximately 30 seconds to make the three responses to each cue. Participants responded verbally to the cues and each response was entered into the computer by the experimenter. Participants were urged to keep the specific instructions in mind (anything, indoors, outdoors) and to try to follow them as closely as possible in generating their responses. Participants were informed that proper nouns were not counted as valid responses, and each of the three responses to a particular cue had to be distinct; that is, second and third responses could not be derivatives of the earlier response(s) for that particular cue (e.g., jog, jogging, jogs). After participants finished all trials in approximately 50 min., they were tested on the N A A R T , debriefed as to the purpose of the experiment, and thanked for their participation. Results and Discussion Participants were screened according to three criteria. Those participants who (1) took substantially longer (i.e., 15 min. longer) than the average 50 min. to complete the experiment, (2) had one-third or more missing responses, and/or (3) had an Estimated Verbal IQ score lower than 95 calculated from the N A A R T were dropped from the analyses. This resulted in excluding six participants, leaving a total of 216 study participants. The frequency of each response to each cue under each of the three instructional conditions was computed for both the word association and the word stem completion tests. Results for the first response were calculated, along with data for responses two and three combined. First responses were analyzed separately from the second and third Intentions 33 responses because of the different constraints that guided the production of first and remaining responses. First responses were constrained by only the words and word stems that were presented as cues and the particular semantic category specified, whereas the production of the later responses was also constrained by the words already produced (see the test instructions). The conditions for the first responses more closely resemble the response conditions in typical word association and stem completion tests, and therefore the data from the first responses give the best estimate of normative performance. In addition, the data for the first responses are more complete than the data for the second and third responses because very few participants failed to come up with at least one response to each cue. The results from both the word association and stem completion tests showed a wide range in the number of different responses produced for each cue. Overall, some cues elicited only 11 different responses, while other cues elicited as many as 127 different . responses. The breakdown of the number of different first responses and second and third responses combined produced per cue for both tasks is shown in Table 1. As would be expected, the second/third responses combined resulted in the production of more varied responses than the first responses alone. Intentions 34 Table 1. Breakdown of the Number of Different Responses Produced for Each Cue in the Word Association Task and the Stem Completion Task Task Type/Instructions Min Max M SD 1 st Response Word Association Indoors 24 57 38.45 7.68 Outdoors 19 56 38.96 7.66 Anything 17 57 35,22 • 8.21 Word Stems Indoors . 11 47 28.48 7.19 Outdoors 17 47 29.80 .6.37 Anything 14 43 • 28.77 ' 6.44 . 2nd and 3rd Responses Combined Word Association -Indoors 56 ' 124 87.57 11.38 Outdoors : 65 119 89.39 11.80 Anything 54 127 85.04 12.05 Word Stems Indoors 25 82 •53.52 11.29 Outdoors 24 84 55.20 "12.62 Anything 25 90 54.59 12,79 .. Intentions 35 Due to the large data set that was collected and space limitations, it is not practical to include the results for every cue in this document. Therefore, to illustrate the range of results, several representative cues (three word cues and three stem cues) with their responses, are shown in Figures 1 through 6. The figures show production frequencies (in terms of percentages) for the 10 most frequent responses for both the first response data and for the second and third response data combined. The ;data for the first response are presented at the top half of the figure and data for the second and third responses combined are presented at the bottom half The data are arranged according to the different instruction conditions (i.e:,indoors vs. outdoors vs. anything). Figures 1 and 2 show cues with at least one response that was strongly biased under the indoor instructions, Figures 3 and 4 are examples o f cues with at least one response biased toward the outdoor instructions, and Figures 5 and 6 are representative of cues that did not lead to any production biases under either the indoor or outdoor instruction., conditions. For all figures the biased words are listed along with the size of the bias in terms of the advantage in production frequency favoring either the indoor or the outdoor subject category. Intentions 36 .Cue: Chair bench indoors dinner Outdoors Figure 1 Example of a word association cue with two responses biased for indoors (i.e., sit, bias = 8.3%; table, bias =11.1%) First responses are shown in the top figure and second and third responses combined are shown in the bottom figure. Intentions 37 C u e : P a Figure 2. Example of a stem completion cue with one response biased for indoors (i.e., paper, bias = 12.5%). First responses are shown in the top figure and second and third responses combined are.shown in the bottom figure. Intentions 38 O H \Afeh penny Outdoors penny Indoors Christmas *; Outdoors Figure 3. Example of a word association cue with one response biased for outdoors (i.e., star, bias = 16.7%). First responses are shown in the top figure and second and third responses combined are shown in the bottom figure. Intentions 39 Cue: RD row(s) Indoors roam Outdoors . Figure 4. Example of a stem completion cue with two responses biased for outdoors (i.e., road, bias = 18.1%; row, bias = 9.8%). First responses are shown in the top figure and second and third responses combined are shown in the bottom figure. Intentions 40 Cue Hand arm Indoors palm Outdoors Figure 5. Example of a word association cue with no biased responses. First responses are shown in the top figure and second and third responses combined are shown in the bottom figure. Intentions 41 Cue: Cb comb Outdoors Figure 6. Example of a stem completion cue with no biased responses. First responses are shown in the top figure and second and third responses combined are shown in the bottom figure. • . Intentions 42 Table 2 shows, for the word associations and word stems separately, the breakdown for the actual number of indoor biased, outdoor biased and neutral (or nonbiased) responses produced. These counts are taken from both the 10 most frequently produced first responses and 10 most frequent second and third responses combined. A response to a particular cue was considered to be an indoor biased response if it had at least an 8% advantage in production frequency over its production frequency in the outdoor condition, and vice versa. A n 8% difference was selected so that the bias would be substantial, but not too substantial that no responses showed this difference. It is clear that the instructional manipulation (i.e., specifying specific semantic categories) had an influence on the production of first responses, but not for second and third responses. That is, for the first responses the particular instructions resulted in some words being produced more frequently in either the indoor or outdoor conditions, depending on which category was specified. Intentions 43 Table 2. Actual Number of Indoor Biased, Outdoor Biased, and Neutral Responses Produced as a Function of test Type and Response Position Category Bias Test Type Indoor Bias Outdoor Bias No Bias Word Associations 1 st Response 2nd + 3rd Responses 35 1 29 1 686 748 Stem Completions 1st Response 2nd + 3rd Responses 48 2 669 745 Intentions 44 Table 2 shows that many more responses were unbiased than biased. This can be largely explained by two factors. First, many of the responses were biased, but did not meet the 8% production frequency bias that was set as the criterion for responses to be considered as biased. Second, each subject had to produce a total of 450 responses, whereas in a typical experiment using these tasks subjects are required to produce about 30 or 40 responses. The results from Experiment 1 show that manipulating the degree of familiarity of retrieval strategies via instructions can influence retrieval from semantic memory, at least for some cues. Specifically, the findings indicate that instructing subjects to generate responses from a particular subject category affects what is retrieved from semantic memory. This is important because it shows that the strategy for searching the. semantic memory domainis not always familiar and automatic. Giving subjects novel criteria for searching their general knowledge can make that search more effortful and influence what is retrieved. . ' . • The data collected from this experiment can serve also as useful normative data for future research. Norms for word stem completions and word associations are presently available in the research literature. For example, Graf and Williams (1987) reported completion norms for.40 three-letter word stems, and Palermo and Jenkins (1964) published word association norms for 200 different stimulus words. However, these data represent general responses to items; specific baseline information for individual items under different instructional/subject category conditions has not been collected. The results of this experiment can be used as a database for selecting target words for the construction of word association and stem completion tests, and perhaps, to guide the. selection of new words for such tests. These data can serve as a good index of baseline performance when different subject categories for retrieval are specified (i.e., indoors and outdoors retrieval . conditions) versus when no constraints or criteria for retrieval are specified (i.e., anything retrieval condition). Experiment 2: Implicit and Explicit Memory Retrieval Intentions Experiment 1 demonstrated that semantic memory retrieval can be influenced by changing the familiarity of retrieval strategies in the absence of a previous study episode. That is, an instructional manipulation specifying particular to-be-searched domains of knowledge (i.e., subject categories) for retrieval affected performance on a task requiring a search of semantic memory. The purpose of Experiment 2 was to examine the importance of the familiarity of search strategies versus the specific retrieval domains targeted by search. In most experiments investigating implicit and explicit memory test performance these two properties of retrieval are confounded. Is there any way to de-confound them? It is not possible to change the specific search domain targeted by implicit and explicit test, but it is possible to manipulate the search strategy familiarity. In a typical explicit memory test, there is a two step strategy process. Subjects must retrieve information from a specific prior episode and then check to see if it meets the criteria for retrieval. By contrast, in a typical implicit memory tests there is only one step, generating responses. Subjects simply generate responses that fit the test cues and do not do any checking. One way to manipulate strategy familiarity in implicit memory tests is to make the retrieval criteria as similar to an explicit test as possible. It may be possible to make implicit recollection more effortful by adding the checking stage to the generation of responses. Thus, Experiment 2 was conducted to test the hypothesis that implicit memory performance can be affected by a shift in the familiarity of the strategy for retrieving information from their respective search domains (i.e., semantic domain vs. episodic domain). To test this hypothesis, memory test instructions specified different subject categories for retrieval which required subjects to check their responses, thereby making performance more effortful.. This was compared to performance when no subject category was specified for retrieval (as in typical implicit tests of memory). Normative data collected from the different task/instruction conditions in Experiment 1 were used as materials. Subjects studied words that were biased in favor of either indoors or outdoors, and their intentions/strategies/criteria for retrieval were manipulated at test. Attest, they were asked to generate/recall words that were either indoor or outdoor words or the first words to Intentions 46 come to mind. Therefore, subjects had to intentionally direct and control their memory search (both implicit and explicit) in order to follow the test instructions for retrieval from different to-be-searched domains. It was hypothesized that memory performance in both the implicit and explicit tests would be highest when study and test conditions matched (i.e., were congruent), and lowest when.study and test conditions were mismatched (i.e., were incongruent). That is, indoor biased targets at study should be produced most often under instructions to generate/recall indoor words, and outdoor biased targets should appear most often under test instructions to generate/recall outdoor words. Targets studied with an indoor bias should be generated/recalled least often, if at all, under outdoor test instructions, and study targets biased for the outdoors should not be produced/recalled in response to indoor retrieval instructions. It was hypothesized that memory performance under the test instructions to • respond with any word that comes to mind would fall somewhere between performance in the congruent and incongruent retrieval conditions. Method Design. The design consisted of a 3 X 2 X 3 X 2 incomplete withinTSubjects design. The factors manipulated were test type (implicit, explicit, baseline), test cue (words or stems), test instructions (indoors, outdoors, or anything), and study word bias (indoors or outdoors). The various experimental conditions are outlined in Table 3. Intentions 47 Table 3. Layout of Design for Experiment 1 Depicting the Different Experimental Conditions. Test Type Experimental Conditions Implicit Test Cue Stems Words Instructions Indoors Outdoors Anything Indoors Outdoors Anything Study bias3 In Out In Out In Out In Out In Out In Out Baseline Test Cue Stems Words Instructions Indoors Outdoors Anything Indoors Outdoors Anything Study bias In Out In Out In .. Out In Out In Out In Out .Explicit • Test Cue Stems Words instructions Indoors Outdoors Indoors Outdoors . Study bias . In Out In Out In Out In Out Study bias refers to whether or not targets were studied in the context of indoors or outdoors. In = indoor bias and Out = outdoor bias. : , Intentions 48 Participants. Participants were 72 student volunteers (52 female and 20 male) from the University of British Columbia who participated in return for credit toward a university psychology course at the 100 or 200 level. The mean level of education of participants was second year university. Materials. The to-be-remembered (TBR) materials were a total of 48 cue-target pairs selected from the normative baseline data collected in the previous experiment. Twenty-four the targets had common words as cues, and 24 had two-letter stems as cues. For the word association test, the cue words were between 3 and 7 letters long ( M = 4.83), and the target words ranged from 3 to 8 letters in length ( M = 4.46). Similarly, for the stem completion test, the target words were between 3 and 7 letters long ( M = 4.75). In order to prevent ceiling and floor effects, and to make the retrieval manipulation as strong as possible, two criteria were followed in the selection of target words from the normative data collected in the previous experiment. First, each target word had to have been produced at least 10% of the time under the anything instruction condition. The lower limit of 10% for production frequency was set to ensure that the targets selected would be produced in response to their respective cues some of the time. Second, there had to be at least an 8% production frequency difference between the indoor and outdoor instructions for each target; an 8% production frequency bias was used in order to select the strongest biased targets possible, and at the same time, obtain the required number of biased targets for the experiment. Therefore, target items selected from the norms for inclusion in the experiment had to be produceable items that were more likely to be used as responses under either the indoor or outdoor instructions. The targets were selected from first.responses only. Refer back to Figures 1 through 4 to see examples of indoor- and outdoor-biased cue-target pairs selected from the norms for inclusion in the experiment. Table 4 shows the breakdown of how each cue-target pair chosen from the norms for use in the experiment compares to the two selection criteria specified above. Shown next to each target is its degree of bias (advantage in production frequency) as a first completion for either the indoor or the outdoor instructions and its frequency of occurrence as a first completion in the anything instruction condition (i.e., baseline). Both frequencies are shown as percentage out of a maximum of 100. Intentions 49 Table 4 Mean Production Frequency of Each Target from its Cue Under Category Biased Instructions and Under Instructions to Produce Any Words that Come to Mind Production Frequency (%) Cue Target Bias 3 (%) Target" (%) Anything0 (%) Outdoor Biased Cues and Responses Word Associations Dogs King Wish Working Black High Loud . Man Foot Moon Salt Soldier Stems D R R A R O SA B E C L KI L A B O PL SC VI cat(s) queen star(s) hard night mountain noise woman shoe(s) sun water war drive rain road sand bee(s) cloud(s) kite(s) lake(s) boys play science vivid 11.1 12.5 . 16.7 : 9.7 16.6 8.3 15.3 11.1 9.7 13.9 . 13.8 15,3 11.1 18.3 18;1 9.7 11.1 9.3 19.4 18 12.5 11.1. 9.8 11.1 38.9 30.6 26.4 1.5.3 . 22.2-12.5 . 37.5 38.9 27.8 18.1 19.4 .27.8 29.2 22.2. 26.4 13.9 15.3 19.5 37.5 20.8 13.9 36.1 41.7 20.8 431 27.8 18.1 18.1 15.3 11.1 29.2 36.1 29.2 8,3 16.7 25 25 9.7 12.5 11.1 18.1 22.2 18.1 6.9 9.7 19.4 31.9 15.3 Intentions 50 Indoor Biased Cues and Responses Word Associations Baby cry 8.4 18.1 13.9 Head ache 11.1 13.9 12.5 Justice peace 12.5 16.7 13.9 Sit down 9,8 29.2 37.5 Fingers nng(s) .8.4 18.1 9.7 Girl friend(s) 8.4 20.9 19.5 Needle thread 22.3 29.2 25 Whiskey drink 11.1 . 22.2 19.4 Afraid dark 31.4 33.3 47.2 Bath room 15.2 19.4 11.1 Chair table 11.1 15.3 12.5 Sleep bed 26.4 36.1 22.2 ems F O food 19.4 33.3 18.1 JO joy • 9.7 33.3 30.6 Q U quiet 9.7 25 23.6 WI windows 11.1 30.6 20.9 B L . block(s) 15.2 29.1 23.6 F A father 25 , 58.3 40.3 IN ' mdoor(s) 27.7 33.3 19.5 M O mother 11.1 20.8 •25' H E help .9.8 18.1 • 15.3 N E neat 11.1 15.3 • . 9.7 OR orange(s) 16.7 • 38.9 23.6 P A paper(s) 12.5 20.8 11.1 ••%Bias is the difference (advantage) in production frequencies between the indoor instruction condition and the outdoor instruction condition. The bias had to be at least 8%. "Target" refers to the production frequency of the target under the specified subject category (i.e., outdoors or indoors). °The anything instruction condition served as a measure of baseline production. Production frequency under this condition had to be at least 10%. Intentions 51 Table 5 provides a summary of the mean production frequency of the targets under baseline and category biased conditions from the normative data. It is clear that the targets were similar in both their baseline production and their production biases under the different category conditions. Intentions 52 Table 5 Production Frequency (%) of Targets from Normative Data Under the Anything and Category Bias Instruction Conditions Category Bias Test Anything Indoors Outdoors Word Association M 21.77 14.68 14.68 SD 10.79 7.75 7.69 Stem Completion • M 19.22 14.92 . 13.29 SD 8.09 6.17 3.91 Overall (words + stems) M 20.49 14.80 13.99 SD 9.44 6.85 . 6.01 Intentions 53 For the study task, the targets from the cue-target pairs were incorporated in sentences that were conceptually biased for either indoors or outdoors, depending on each target's original bias. That is, targets selected from the norms because they were biased in the direction of indoors were biased to an even greater extent by being placed in a sentence that emphasized the indoors. Conversely, targets selected from the norms on the basis of their outdoor bias, were further biased by being placed in a sentence that highlighted the outdoors. A total of 48 sentences were constructed, one for each target word, 24 of which were indoor biased sentences and 24 outdoor biased sentences. The sentences are listed in Appendix A. Each sentence was.shown on 4" X 6" index cards in 12 pt. Times New Roman font. The target word in each sentence was printed in bold and underlined text for emphasis. The sentences were constructed in such a way that no words were repeated (with the exception of pronouns, conjunctions, etc.) in and between sentences. The degree to which the sentences were biased indoors or outdoors was judged independently by three graduate students. Sentences in which the intended bias was unclear were revised until all three judges were in agreement of the direction of the bias. For counterbalancing purposes, the 48 cue-target pairs were divided randomly into two sets of 24. Each participant was shown one of two sets of targets, in random order, at study and was later tested with both word association and stem completion implicit tests, followed by an explicit cued recall test. The other critical set of cue-target pairs was not presented at study. The nonpresented set provided an estimate of baseline performance. Three different test response forms were constructed to aid counterbalancing (see Appendix B for a sample response form). Each test form consisted of seven pages, with one page for study, one for practice, three for.the implicit tests (indoors vs. outdoors vs. anything), and two for the explicit test (indoors vs. outdoors). Three pages were required for the implicit portion, one for each category instruction condition, and two pages were needed for the explicit test,' with one used for each instruction condition. Test items were counterbalanced with test instructions, so that across test forms all cues occurred in each retrieval instruction condition. Within each retrieval instruction condition, the cues were ordered in a random fashion. The experiment was arranged such that the two study target sets and the three response forms were completely counterbalanced. Intentions 54 Tests; Participants in this experiment completed both implicit and explicit tests of memory. Two implicit tests were used: (1) word association and (2) stem completion. For the word association test, participants were presented with word cues (e.g., Mountain: ), whereas for the stem completion, two-letter word stems were presented as cues (e.g., ST ). The explicit test of memory consisted of a cued recall test which used cues identical to those used in the implicit tests (words and stems), and thus, only the instructions differed between the two tests. Procedure. The general procedure consisted of instruction, study, practice, implicit testing, then explicit testing. Participants were tested individually. They were given a consent sheet to read, and following their written consent to take part in the experiment, they were handed a response sheet (see Appendix B) and were asked to read the directions. The experiment was described as examiningword generation with the help of different categories guided by different instructions (i.e., indoors, versus outdoors, versus anything). The study task was introduced as a task designed to familiarize participants with the indoor/outdoor distinction. During the study phase, participants were shown 24 sentences and were asked to rate each sentence according to whether they thought it described something or a situation occurring primarily indoors or outdoors. A 5-point rating scale was used, with T representing 'completely indoors' through to '5' representing 'completely outdoors'. In each sentence there was a bolded, underlined word (the target word). Participants were asked to say that word aloud in order to. focus their attention, then read the sentence to themselves and rate it according to the scale. The real purpose of getting participants to say the target word aloud was to ensure encoding. The sentences containing the target words were presented in a random order for each participant at the rate of approximately 3 seconds per sentence. Immediately after study, practice trials (see Appendix B) were administered to familiarize participants with the stem completion and word association tasks. For each practice trial, subjects generated a word for one stem and for one word cue according to each instruction condition. If participants had difficulty generating responses to any of the practice cues, the experimenter provided some examples of acceptable responses. Intentions 55 Participants were advised that proper nouns were not valid responses and that only one word responses were acceptable. For the implicit test phase, participants were asked to generate words for each cue , (words or stems), as quickly as possible, following the appropriate category specified for retrieval. The cues were organized in three blocks of 16, with retrieval instructions (ie., indoors, outdoors, anything) written directly above each block (see Appendix B). The order of the instruction depended on the response form used. The importance of following the particular instructions for each block as closely as possible was emphasized by reminding subjects to keep the specified category in mind when generating responses to the cues. After finishing one block of cues, the participants read the instructions for the next block, completed that block, and continued until responses were provided for all cues in all blocks. After completion of the implicit test phase, participants were introduced to the explicit cued recall test. Participants were asked to think back to the sentences they had studied, and specifically to the bolded, underlined words they had spoken aloud (i.e., the target words) from these sentences. The test had two blocks of cues, one for the indoor instruction condition and one for the outdoor instruction condition. These were the same blocks of cues that were used for the implicit test (see Appendix B). Participants were instructed to use the cues to try to remember the target words from the sentences that occurred in either the indoor sentences or the outdoor sentences. Participants were told that some of the test cues did not correspond to targets on the previous study list and that not all of the cues would necessarily lead to a correct response; and were instructed to write answers for those cues that lead to responses congruent with the specific instructions. After participants completed both memory tests they were debriefed as to the purpose of the experiment (see Appendix C) and thanked for their participation. The experiment required about 30 minutes. Intentions 56 Results and Discussion The main dependent measures were the number (i.e., the count) of target words remembered on the word association and stem completion tests, and the number of targets produced on the cued recall test. For both tests, a target word was scored as correct only if it was produced in response to the appropriate cue. Plurals, word derivatives (e.g., for jog, jogging; for mother, mom), and unambiguous misspellings were counted as valid responses. The significance level for all statistical tests was set at .05. Word association and stem completion. Analyses for the implicit tests were conducted separately on the data from the word association test and from the stem completion test. To allow for comparisons of performance across conditions (i.e., implicit and baseline), performance in each test condition was calculated in terms of the proportion of targets produced out of the total possible target productions when possible. When performance could not be calculated in terms of proportions, the actual, number of targets produced was used instead. Overall, the pattern of results shows that, as hypothesized, priming was higher when study and test conditions were congruent (i.e., indoor-indoor, outdoor-outdoor) than when they were incongruent (i.e., indoor-outdoor, outdoor-indoor). Clearest evidence for this finding was demonstrated in the stem completion test. Figures 7 and 8 show the mean proportions of cues completed with study-list targets in the main experimental conditions. Intentions 57 Implicit Baseline Indoors Outdoors Anything Indoors Outdoors Anything Test Instructions Indoor Biased Targets Outdoor Biased Targets Figure 7. Mean proportion of targets produced under each target bias/test instruction condition on the word association test. (Proportion = number of targets produced / total number of targets). Intentions 58 0.5 Indoors Outdoors Anything • Indoors Outdoors Anything Test Instructions Indoor Biased Targets Outdoor Biased Targets Figure 8. Mean proportion of targets produced under each target bias/test instruction condition on the stem completion test. (Proportion = number of targets produced / total number of targets). Intentions 59 Priming refers to the advantage previously studied, or primed, items have over new, unstudied items on implicit tests of memory. There are two predominant ways in which priming has been measured: in terms of difference scores and in terms of relative priming scores. A difference measure of priming is simply the difference between performance on studied items minus unstudied items, whereas one relative measure of priming is the difference between studied minus unstudied items divided by the score for unstudied items (i.e., baseline performance) (see Snodgrass, 1989a, 1989b). Another way to compute relative priming is to divide the difference score by 1 - the corresponding baseline score. These two ways of calculating relative priming scores will be referred to as relative bottom and relative top priming scores, respectively. The difference priming scores and the two relative priming scores for the data in all cue bias/instruction conditions are shown in Table 6. Intentions 60 Table 6. Priming Measured in Terms of Difference Scores and Relative Scores For the Different Experimental Conditions Priming Baseline Difference3 Relative top b Relative bottom(%)c M SD Word Associations Indoor Biased Targets Indoor Instructions (.19) (.22) .06 .07 . 31.58 Outdoor Instructions (03) (.09) .03 .03 100.0 Anything Instructions (.17) (.18) .03 .04 17.65 Outdoor Biased Targets Indoorlnstructions (.06) (.11) .04 .04 66.67 Outdoor Instructions (19) (14) 0 0 0 Anything Instructions (.26) •(•24) , .05 .07 19.23 Word Stems Indoor Biased Targets Indoor Instructions (.16) (.22) 19 ..23 118.75 Outdoor Instructions (.08) (14) .02 .02 75.0 Anything Instructions (11) (.16) .14 ' 16 , 127:27 Outdoor Biased Targets Indoor Instructions (.01) (.06) • 05 .05 500.0 , Outdoor Instructions (.26) (23) .16 . .22 61.54 Anything Instructions (.15) (14) .11 . 13 73.33 'Difference priming score = no. of previously studied targets produced minus baseline performance (i.e., no. targets produced in the absence of study). b Relative priming score = difference priming score / (1 minus mean baseline score for corresponding condition). 0 Relative priming score = difference priming score / corresponding baseline score. Intentions 61 Most priming studies (e.g., Challis & Brodbeck, 1992; Jacoby & Dallas, 1981; Roediger, Weldon, Stadler, & Reigler, 1992; Weldon, 1993) use the difference between studied and unstudied items to measure priming effects. This measurement of priming is most applicable when baseline performance across conditions is comparable. However, when baseline varies significantly it should be corrected for to permit comparisons across tasks The baseline performance of the two implicit tests in this experiment differed over a range of .01 to .26. Relative priming measures provide an estimate of the amount of priming relative to the given baseline level of performance. However, even relative priming measures can be a problem when performance in some conditions is, or when baseline is, on the floor (e.g., .01 and .03). It is apparent from Table 6 that different patterns of results can be found depending on how priming is measured. Overall, it can be seen from looking at Table 6 that priming scores were larger for stem completion than for word association, regardless of which measurement of priming was used. When the three different measurements of priming are considered separately, it is apparent that there are problems when priming is calculated using difference scores and relative bottom scores (see Table 6). These problems arise primarily from the fact that baseline performance is very low in several conditions (i.e., .01, .03, .06). Surprisingly, even though these baseline scores are small, all are significantly off the floor except for. outdoor biased targets with indoor instructions for the word stems (t(35) = 1.43, p=. 160) Nevertheless, low baseline performance renders relative bottom scores unusable because . unusual priming scores result (e.g., 500% when the difference score was only .05). When priming, is measured in terms of difference scores, none of the priming scores for the word association test are significantly different from baseline. However, priming is significant in all conditions for the word stem completion test except for indoor biased targets with outdoor instructions (t(35) = .94, p_ = 35) and outdoor biased targets with indoor instructions (t(35) = 1.53, p = . 14). Although there is significant priming in the stem completion test, conducting analyses using difference scores would be problematic because baseline performance in the different conditions is so varied. As stated previously, difference priming scores are only meaningful when baseline scores are comparable. Intentions 62 Therefore, the only priming scores that can be relied on to analyze are the relative top scores. It was hypothesized that priming scores would be larger when study and test conditions were congruent (i.e., indoors-indoors, outdoors-outdoors), whereas priming scores would be smallest when study and test conditions were incongruent (i.e., indoors-outdoors, outdoors-indoors), and performance in the anything test condition would fall somewhere in between. In order to do the analyses the data were combined into three different groups according to cue-target bias and category retrieval instructions: congruent, incongruent, and anything. When relative (top) priming scores (i.e., difference score / 1 - mean baseline for corresponding condition) were used to measure priming the pattern of scores for the stem completion test was consistent with the hypotheses (see Table 6). Significant priming for stem completion was found for the congruent and anything conditions (t(71) = 3.54, p_ = -.001 and t(71) = 4.79, p = .001 respectively), but not for the incongruent condition. A repeated-measures A N O V A showed a main effect for test condition, F(2,142) = 5.37, M S R = 85. Follow up paired-samples t-tests indicated that priming in both the congruent, and anything conditions was significantly greater than in the incongruent condition (t(71) = 2.79, p = .007 and t(71) = -2.87, p = .005 respectively), but the two conditions did not differ significantly from one another. The data from the word association test were not analyzed because there were no significant priming effects. In general, the results indicate that, as hypothesized, priming (as measured by the relative top scores) was higher when study and test conditions were the same (i.e., congruent) than when they were different (i.e., incongruent). Priming in the anything condition also was greater than in the incongruent condition, however, the anything and congruent conditions did not differ significantly. This pattern of results was found for the stem completion test, but not for the word association test. It is important to note that, although problematic because of varied baseline scores, the difference priming scores for the stem completion test show the same general pattern of scores. In addition, although not significant, the relative top priming scores for the word associations also show a pattern somewhat like that of the stem completions. . Intentions 63 Cued recall. Figure 9 shows the results for the cued recall test. Although cued recall was quite low, it was significantly off the floor in all conditions, minimum t(71) = 3.56, p = .001. A repeated-measures A N O V A showed a two-way test cue (words vs. stems) by task type (congruent vs. incongruent) interaction, F ( l , 71) = 5.10, M S e = .46. A further analysis showed that the congruent and incongruent test conditions were significantly different for both test cues. For the word cues^ F ( l , 71) = 15.74, M S K = .35, the congruent test conditions ( M = .60) showed significantly better recall than the incongruent conditions ( M = 21). Similarly, the stem cues resulted in better recall when the test conditions were congruent ( M = .94) rather than incongruent ( M = .19), F ( l , 71) = 38.09, M S e = 5 3 . Thus, cued recall was significantly better when study and test conditions were congruent than when they were incongruent. Intentions 64 Congruent Incongruent Words Stems Type of Cue Figure 9. Mean number of words recalled for the word cues and stem cues under the different cue bias/test conditions (i.e., congruent and incongruent) for the explicit cued recall test. Intentions 65 General Discussion The purpose of this thesis was to gain further insight into implicit and explicit memory test performance. Both implicit and explicit recollection are initiated and guided by conscious, intentional retrieval. The intentional retrieval in implicit memory tests differs from the intentional retrieval in explicit memory tests in terms of the to-be-searched memory domain (i.e., semantic memory domain vs. episodic memory domain) and in terms of the degree of familiarity of the search strategies (i.e., familiar, fluent vs. novel, effortful). In most previous studies investigating implicit and explicit memory phenomena, the search domain and the familiarity of the search strategy have been confounded. The present experiments were an attempt to tease apart the effects of these two aspects of memory retrieval to gain a better understanding of which (or both) is crucial for implicit and explicit memory performance. The results revealed that differences in the familiarity of search strategies can affect what is retrieved from memory. Experiment 1 demonstrated that a manipulation of the familiarity of a strategy for retrieval from semantic memory, in the absence of a prior study episode, can influence retrieval. That is, semantic memory performance can be shifted around according to test instructions that specify either a familiar search or a novel search of different categories or domains from which to generate responses. More importantly, Experiment 2 demonstrated similar findings by showing that changes in the familiarity of a search strategy can influence semantic memory retrieval following a study episode (i.e., implicit memory). This effect was found for a word stem completion test, but not for a word association test. Explicit cued recall also revealed an influence due to the search strategy familiarity manipulation, with both word cues and stem cues showing significant effects. These results represent an important first attempt at separating the effects of search domain and search strategy familiarity on memory performance. The data from the stem completions provide some evidence that it may not be the familiarity of the search strategy that is important for implicit and explicit memory performance, but the particular search category or domain at which retrieval is targeted. However, these results are preliminary and must be interpreted with caution. One reason for cautious interpretation is that baseline Intentions 66 performance was too low in several of the experimental conditions and was quite variable. This made the calculation of priming scores difficult. Second, different patterns of priming resulted depending on which method for measuring priming was used. However, it is important to note that for the stem completions the relative top priming scores and the difference scores showed a similar pattern of priming (see Table 6). Third, the influence of manipulating the familiarity of the search strategy in the implicit memory tests was found for the word stem completion test but not for the word association test. It is not clear as to why priming was not significant when word associations were used, but it is possible that the manipulation employed was not strong enough. New efforts are currently being made to come up with different category distinctions for retrieval other than the indoor/outdoor distinction that was used in these experiments. It may have been that the indoor and outdoor categories overlapped too much to permit a strong enough manipulation of search strategy familiarity. The present experiments utilized one methodology in an attempt to de-confound search strategy familiarity and search domain in implicit recollection. Other methods may result in a more effective manipulation of the familiarity/fluency of the search strategy. For instance, attention manipulations have often been used to assess whether or not a cognitive process is effortful, and have recently emerged in implicit and explicit memory research (e.g., Mulligan & Hartman, 1996; Parkin & Russo, 1990; Parkin, Reid, & Russo, 1990; Weldon & Jackson-Barrett, 1993). It is possible to determine whether or not memory retrieval is effortful and attention-demanding by dividing subjects' attention between the memory task and another potentially distracting task. The results of the recent studies that have employed a divided attention task, indicate that explicit memory performance is diminished when attention is divided at study, whereas the effect on implicit memory is less clear. It appears that implicit memory performance can be affected or not affected by divided attention at study depending on variables such as the exposure time of items at study (e.g., Weldon & Jackson-Barrett, 1993) and the type of implicit memory test used (i.e., tests that focus more of the conceptual or the perceptual properties of stimuli) (e.g., Mulligan & Hartman, 1996). Intentions 67 Although the effects of dividing attention at study on implicit memory performance are not clear, examining its effects during memory search could be revealing. In principle, memory search strategies could be made more controlled and effortful if an attention-demanding retrieval task was employed. By making implicit memory tasks more attention demanding and effortful, the importance of the familiarity of the search strategy on implicit memory performance could be investigated further. The present research was intended as a preliminary examination of the effects search domain and search strategy familiarity on implicit and explicit memory test performance. The results from the stem completion data suggest that implicit memory retrieval can be guided by an effortful, demanding search, and therefore is not always a familiar, fluent process. The results, although preliminary, are promising. They suggest that the particular search category or domain targeted by retrieval may be more critical than the familiarity of the search strategy in implicit and explicit memory test performance. The challenge for future work is to develop new materials and alternative ways for testing the differential effects of search domain and search strategy familiarity on implicit and explicit recollection. The potential implications for these findings are considerable. For example, new insight could be gained into the selective memory deficits observed in amnesic patients. As discussed previously, amnesic patients show a major deficit in performance on explicit tests of memory, but demonstrate little or no impairment in implicit memory test performance. If further research confirms the present findings, it could be argued that it is not the degree of fluency/automaticity of the retrieval search that determines which memory abilities suffer and which are preserved in amnesic patients, but the specific memory domain at which the search is targeted. These results also could potentially help explain other findings such as why explicit recollection, but not implicit recollection, changes with age and why implicit memory abilities develop and become fully functional at an earlier age than explicit memory abilities. Intentions 68 References Bassili, J. N . , Smith, M . C , & MacLeod, C. M . (1989). Auditory and visual word stem . completion: Separating data driven and conceptually-driven processes. Quarterly Journal of Experimental Psychology: 41(A), 439-445. Blaxton, T. A. (1989). Investigating dissociations among memory measures: Support for a transfer appropriate processing framework. Journal of Experimental Psychology: Learning. Memory, and Cognition. 15, 657-668. Bransford, J. D. , Franks, J. J. , Morris, C D . , & Stein, B. S, (1979). Some general constraints on learning and memory research. In L : S. Cermak, & F. I. M . Craik (Eds), Levels of processing in human memory (pp. 331-354). Hillsdale, NJ: Lawrence Erlbaum. Cermak, L . S. (1993). Automatic versus controlled processing and the implicit task performance of amnesic patients. In P. Graf, & M E . J . Masson (Eds), Implicit memory: New directions in cognition, development, and neuropsychology (pp. 287-301). Hillsdale, NJ: Lawrence Erlbaum. Challis, B. H . , & Brodbeck, D . R. (1992). Level of processing affects priming in word fragment completion. Journal of Experimental Psychology: Learning, Memory, and Cognition. 18. 595-607. Clarke, R , & Morton, J. (1983). Cross modality facilitation in tachistoscopic word recognition. Quarterly Journal of Experimental Psychology, 3 5 A: 79-76. Cohen, N , & Squire, L . R. (1980). Preserved learning and retention of pattern analyzing skills in amnesia: Dissociation of knowing how and knowing that. Science, 210, 207-210. ; , • ... . Craik, F. I. M . (1983). On the transfer of information from temporary to permanent memory. Philosophical Transactions of the Royal Society, London, 302B, 341-359. Craik, F. I. M . , & Byrd, M . (1982). Aging and cognitive deficits: The role of attentional resources. In F. I. M . Craik, & S. Trehub (Eds), Aging and cognitive processes (pp. 191-211). New York: Plenum. Intentions 69 Craik, F. I. M . , & Jennings, J. M . (1992). Human memory. In F. I. M . Craik, & T. A. Salthouse (Eds), The handbook of aging and cognition (pp. 51-110). Hillsdale, NJ: Lawrence Erlbaum. Craik, F. I. M . , Moscovitch, M . , & McDowd, J. M . (1994). Contributions of surface and conceptual information to performance on implicit and explicit memory tasks. Journal of Experimental Psychology: Learning, Memory, and Cognition, 20, 864-. 875. Craik, F. I. M . , & Tulving, E . (1975). Depth of processing and the retention of words in episodic memory. Journal of Experimental Psychology: General, 104, 268-294. Dunn, J. C , & Kirsner, K. (1988). Discovering functionally mental processes: The principle of reversed association. Psychological Review, 95, 91-101. Dunn, J. C , & Kirsner, K. (1989). Implicit memory: Task or process? In S. Lewandowsky, J. C. Dunn, & K. Kirsner (Eds), Implicit memory: Theoretical issues (pp. 17-31). Hillsdale, NJ: Lawrence Erlbaum. Durgunoglu, A . Y . , & Roediger, H . L . (1987). Test differences in accessing bilingual memory. Journal of Memory and Language, 26, 377-391 Graf, P. (1990). Life span changes in implicit and explicit memory. Bulletin of the Psychonomic Society, 28, 353-358. Graf,'P. (1991). Implicit and explicit memory: A n old model for new findings. In W. Kessen, A. Ortony, & F. Craik (Eds), Memories, thoughts, and emotions: Essays in honor of George Mandler (pp. 135-147). Hillsdale, NJ: Lawrence Erlbaum.. Graf, P., & Birt, A . R. (1996). Explicit and implicit memory retrieval: Intentions and strategies. In L . Reder (Ed) , Implicit memory and metacognition (pp. 25-44). Mahwah, NJ: Lawrence Erlbaum. Graf, P., & Komatsu, S. (1994). Process dissociation procedure: Handle with caution. European Journal of Cognitive Psychology, 6, 113-129. Graf, P., & Mandler, G. (1984). Activation makes words more.accessible but not necessarily more retrievable. Journal of Verbal Learning and Verbal Behavior, 23, 553-568. Intentions 70 Graf, P., & Ryan, L . (1990). Transfer-appropriate processing for implicit and explicit memory. Journal of Experimental Psychology: Learning, Memory, and Cognition, J A 978-992. Graf, P., & Schacter, D L . (1985). Implicit and explicit memory for new associations in normal and amnesic participants. Journal of Experimental Psychology: Learning, Memory, and Cognition. 11. 501-518. Graf, P., & Schacter, D. L . (1987). Selective effects of interference on implicit and explicit memory for new associations Journal of Experimental Psychology: Learning, Memory, and Cognition, 13, 45-53. Graf, P., & Schacter, D. L . (1989). Unitization and grouping mediate dissociations in memory for new associations Journal of Experimental Psychology Learning, Memory, and Cognition. 15, 930-940. Graf, P., Shimamura, A . P., & Squire L . R. (1985). Priming across modalities and priming across category levels: Extending the domain of preserved functioning in amnesia. Journal of Experimental Psychology: Learning, Memory, and Cognition. 10, 385-395. Graf, P., Squire, L . R , & Mandler, G. (1984). The information that amnesic patients do not forget Journal of Experimental Psychology: Learning, Memory, and Cognition. 10. 164-178. Graf, P., & Williams, D. (1987). Completion norms for 40 three-letter word stems. Behavior Research Methods. Instruments, & Computers, 19, 422-445. Greenbaum, J. L . , & Graf, P. (1989) Preschool period development of implicit and explicit remembering. Bulletin of the Psychonomic Society, 27. 417-420. Hamann, S. B. (1990). Level-of-processing effects in conceptually driven implicit tasks. Journal of Experimental Psychology: Learning, Memory, and Cognition, 16, 970-977. Hasher, L . , & Zacks, R. T. (1979) Automatic and effortful processes in memory. Journal of Experimental Psychology: General, 108, 356-388. Hirst, W. (1982). The amnesic syndrome: Descriptions and explanations. Psychological Bulletin. 91. 435-460. Intentions 71 Howard, D. V . , Heisey, J. G , & Shaw, R. J. (1986). Aging and the priming of newly learned associations. Developmental Psychology, 22, 78-85. Howes, J. L . , & Katz, A . N . (1992). Remote memory: Recalling autobiographical and public events from across the lifespan Canadian Journal of Psychology, 46, 92-116. Humphreys, M . S., Bain, J. D. , & Pike, R. (1989). Different ways to cue a coherent memory system. A theory for episodic, semantic, and procedural tasks. Psychological Review. 96. 208-233. Hunt, R. R., & Toth, J. P. (1990). Perceptual identification, fragment completion, and free recall: Concepts and data. Journal of Experimental Psychology: Learning. Memory, and Cognition. 16. 282-290. Jacoby, L . L . (1983). Remembering the data: Analyzing interactive processes in reading. Journal of Verbal Learning and Verbal Behavior, 22, 485-508. Jacoby, L . L . (1991). A process dissociation framework: Separating automatic from intentional uses of memory. Journal of Memory and Language, 30, 513-541. Jacoby, L . L . (1993). Redefining automaticity: Unconscious influences, awareness, and control: In A. Baddeley, & L . Weiskrantz (Eds), Attention: Selection, awareness. and control (pp. 261-282). Oxford, England: Clarendon Press. Jacoby, L . L . (1994). Measuring recollection: Strategic versus automatic influences of associative context. In C. Umilta, & Moscovitch, M . (Eds), Attention and performance X V : Conscious and nonconscious information processing (pp. 661-679). Cambridge, M A : Bradford. Jacoby, L . L . , & Dallas, M . (1981). On the relationship between autobiographical memory and perceptual learning Journal of Experimental Psychology: General, 110, 306-340. Jacoby, L . L . , Ste-Marie, D , & Toth, J. P. (1993). Redefining automaticity: Unconscious influences, awareness, and control. In A. Baddeley, & L . Weiskrantz (Eds), Attention: Selection, awareness, and control (pp. 261-282). Oxford, England: Clarendon Press. James, W. (1962). Psychology. Briefer course. New York. Collier. (Original work published 1890). Intentions 12 Jennings, J. M . , & Jacoby, L . L . (1993). Automatic and intentional uses ofmemory: . Aging, attention, and control. Psychology and Aging, 8, 283-293. Kirsner, K . , & Smith, M . C. (1974). Modality effects in word identification. Memory and Cognition. 2. 637-640 Kolers, P. A . (1975a). Memorial consequences of automatized encoding. Journal of Experimental Psychology: Human Learning and Memory. 1, 689-701. Kolers, P. A. (1975b). Specificity of operations in sentence recognition. Cognitive Psychology. 7. 289-306. •'''*'. Kolers, P: A . (1976). Reading a year later. Journal of Experimental Psychology. Human Learning and Memory. 2, 554-565. Kolers, P. A . (1979). A pattern-analyzing basis of recognition I n L . S. Cermak, & F . I. M . Craik (Eds), Levels of processing in human memory (pp. 363^384). Hillsdale, NJ:' 1 Lawrence Erlbaum. _ Kolers, P A . , & Roediger, H . L . (1984). Procedures of mind. Journal of Verbal Learning . . and Verbal Behavior, 23. 425-449. Komatsu, S., Graf, P., & Uttl, B . (1995). Process dissociation procedure: Core assumptions fail, sometimes. European Journal of Cognitive Psychology. 7. 19-40. Light, L . L „ & Singh, A . (1987). Implicit and explicit memory in young and older adults. Journal of Experimental Psychology: Learning, Memory, and Cognition, 13, 531-. 541. , Lockhart, R. S, (1989). The role of theory in understanding implicit memory. In S. Lewandowsky, J. C. Dunn. & Kirsner. K. (Eds.). Implicit memory: Theoretical issues (pp, 3-13). Hillsdale, NJ: Lawrence Erlbaum. , McAndrewS; M . P., Glisky, E . L . , & Schacter, D L , (1987). When priming persists: Long-lasting implicit memory for a single episode in amnesic patients! Neuropsychologica, 25. 497-506. McKoon, G . , Ratcliff, R., & Dell, G . S, (1986) A critical evaluation of the semantic-episodic distinction. Journal of Experimental Psychology: Learning, Memory, and Cognition. 12: 295-306. Intentions 73 Morris, C. D . , Bransford, J. D . , & Franks, J. J. (1977). Levels of processing versus transfer appropriate processing. Journal of Verbal Learning and Verbal Behavior, 16, 519-533. Mulligan, N . W., & Hartman, M . H . (1996). Divided attention and indirect memory tests. Memory and Cognition, 24, 453-465. Nadel, L . (1992). Multiple memory systems: What and why? Journal of Cognitive Neuroscience. 4, 179-188. Neely, J. H . (1989). Experimental dissociations and the episodic/semantic memory distinction. In H . L . Roediger, & F. I. M . Craik (Eds), Varieties of memory and consciousness. Essays in honour of Endel Tulving (pp. 229-270). Hillsdale, NJ. Lawrence Erlbaum. Neill, W T., Beck, J. L . , Bottalico, K S., & Molloy, R D (1990). Effects of intentional versus incidental learning on explicit and implicit tests of memory. Journal of Experimental Psychology: Learning, Memory, and Cognition. 16, 457-463. Nelson, D. L . (1989). Implicitly activated knowledge and memory. In C. Izawa (Ed) , Current issues in cognitive psychology: The Tulane Flowerree symposium on cognition. New York. N Y : Lawrence Erlbaum. Nelson, D. L . (1994). Implicit memory In P. E . Morris, & M . Gruneberg (Eds), Theoretical aspects of memory (2nd ed.) (pp. 130-167). New York, N Y . Routledge. Nelson, H : E . (1982). National Adult Reading Test (NART): Test Manual. Windsor, U K . N F E R Nelson. Nelson, H . E . , & O'Connell, A. (1978). Dementia: The estimation of pre-morbid intelligence levels using the new adult reading test. Cortex, 14, 234-244. Palermo, D. S , & Jenkins, J. J. (1964). Word association norms: Grade school through college. Minneapolis: University of Minnesota Press. Parkin, A . J. (1982). Residual learning capability in organic amnesia. Cortex, 18, 417-440. Parkin, A. J. (1987). Memory and amnesia. Oxford, England: Blackwell. Intentions 74 Parkin, A. J. (1993). Implicit memory across the lifespan. In P. Graf, & M . E. J. Masson (Eds), Implicit memory; New directions in cognition, development, and neuropsychology (pp. 191-206). Hillsdale, NJ: Lawrence Erlbaum. Parkin, A. J., Reid, T. K., & Russo, R. (1990). On the differential nature of implicit and explicit memory. Memory and Cognition, 18, 507-514. Parkin, A. J., Russo, R. (1990). Implicit and explicit memory and the automatic/effortful distinction. European Journal of Cognitive Psychology, 2, 71-80. Posner, M . I , & Snyder, C. R. R. (1975). Attention and cognitive control. InR. L. Solso (Ed), Information processing in cognition: The Loyola Symposium (pp. 58-85). ' Hillsdale, NJ: Lawrence Erlbaum. Raaijmaker, J. G. W , & Shiffrin, A . - M . (1981). Search of.associative memory. Psychological Review. 88. 93-134. ; Richardson-Klavehn, A., & Bjork, R. A. (1988). Measures of memory. Annual Review of Psychology. 39, 475-543. Roediger, H. L. (1990a). Implicit memory: Retention without awareness. American Psychologist, 45, 1043-1056. Roediger, H. L. (1990b) Implicit memory: A commentary. Bulletin of the Psychonomic Society. 28, 373-380. Roediger, H. L , & Blaxton, T. A. (1987a). Effects of varying modality, surface features, and retention interval on priming in word-fragment completion. Memory and Cognition. 15, 379-388. Roediger, H. L. , & Blaxton, T. A. (1987b). Retrieval modes produce dissociations in memory for surface information. In D. S. Gorfein & R. R. Hoffman (Eds), Memory and cognitive processes: The Ebbinghaus Centennial Conference (pp. 349-379). Hillsdale, NJ: Erlbaum. Roediger, H. L. , & McDermott, K. B. (1993). Implicit memory in normal human participants. In F. Boiler & J. Grafman (Eds), Handbook of neuropsychology, (Vol. 8, pp. 63-131). New York, N Y : Elsevier. Roediger, H. L. , & Rajaram, S. (1990). Specifying criteria for postulating memory systems. Conference on the National Institute of Mental Health et al.: The development and Intentions 75 neuralAbases of higher cognitive functions (1989,Philadelphia, Pennsylvania). , Annals of the New York Academy of Sciences. 608, 572-595. Roediger, H . L , & Srinivas, K. (1993). Specificity of operations in perceptual priming. In P. Graf, & M E . J. Masson (Eds), Implicit memory. New directions in cognition. development, and neuropsychology (17-48). Hillsdale, NJ: Lawrence Erlbaum. Roediger, H . L . , & Weldon, M . S. (1987). Reversing the picture superiority effect. In M A. McDaniel, & M . Pressley (Eds), Imagery and related mnemonic processes: Theories; individual differences, and applications (pp. 151-174). New York, N Y : Springer-Verlag. Roediger, H . L . , Weldon, M S . , & Challis, B. H . (1989). Explaining dissociations between implicit and explicit measures of retention: A processing account. In H . L . Roediger, & F. I. M . Craik (Eds). Varieties of memory and consciousness: Essays in honour of Endel Tulving (pp. 3-14). Hillsdale, NJ: Lawrence Erlbaum. Roediger, H . L . , Weldon, M . S., Stadler, M . L . , & Riegler, G . L . (1992). Direct comparison of two implicit memory tests: Word fragment and word stem completion. Journal of Experimental Psychology: Learning, Memory, and Cognition. 18. 1251-1269. Schacter, D. L . (1987). Implicit memory: History and current status. Journal of Experimental Psychology: Learning, Memory, and Cognition, 13, 501-518. Schacter, D. L . (1989). On the relation between memory and consciousness: Dissociable interactions and conscious experience. In H L Roediger, & F. I. M . Craik (Eds), Varieties of memory and consciousness: Essays in honour of Endel Tulving (pp. 355-389): Hillsdale, NJ: Erlbaum. Schacter, D. L . (1990). Perceptual representation systems and implicit memory: Toward a, resolution of the multiple memory systems debate. Annals of the New York Academy of Sciences, 608, 543-571 Schacter, D. L . (1992). Priming and multiple memory systems: Perceptual mechanisms of implicit memory. Journal of Cognitive Neuroscience, 4, 244-256 Schacter, D. L . , Bowers, J., & Booker, J. (1989). Intention, awareness, and implicit . memory: The retrieval intentionality criterion. In S. Lewandowsky, J. C. Dunn, & Intentions 76 K. Kirsner (Eds), Implicit Memory: Theoretical Issues (pp. 47-65). Hillsdale, NJ: Lawrence Erlbaum. Schacter, D. L . , & Graf P. (1986). Effects of elaborative processing on implicit and explicit memory for new associations. Journal of Experimental Psychology: Learning, Memory, and Cognition. 12, 432-444. Schacter, D. L : , & Graf, P.(1989). Modality specificity o f implicit memory for new associations. Journal of Experimental Psychology: Learning. Memory, and Cognition. 15. 3-12. . Schacter, D. L . , & McGlynn, S. M . (1989). Implicit memory: Effects of elaboration depends on unitization. American Journal of Psychology. 102. 151-181. Schacter, D . L . , & Tulving, E . (1982). Memory, amnesia, and the episodic/semantic distinction. In R. L . Isaacson, & N . E . Spear (Eds), The expression of knowledge (pp. 33-65). New York: Plenum Press. Seidenberg, M . S., & McClelland, J. L . (1989). A distributed, developmental model of word recognition and naming. Psychological Review. 96, 523-568. Sherry, D . F., & Schacter, D. L . (1987). The evolution of multiple memory systems. Psychological Review, 94, 439-454. Shiffrin, R. M . (1970). Memory search. In D. A . Norman (Ed) , Models of human memory. New York, N Y : Academic Press. Shimamura, A. P. (1986). Priming effects in amnesia: Evidence for a dissociable memory function. Quarterly Journal of Experimental Psychology, 38A, 619-644. Shimamura, A. P. (1993). Neuropsychological analyses of implicit memory: History, methodology and theoretical interpretations. In P. Graf, & M . E . J. Masson (Eds.), Implicit memory: New directions in cognition, development, and neuropsychology (pp. 265-285). Hillsdale, NJ: Lawrence Erlbaum. Snodgrass, J. G. (1989a). How many memory systems are there really?: Some evidence from the picture fragment completion task. In C. Izawa (Ed) , Current issues in cognitive processes (pp. 135-173). Hillsdale, NJ: Lawrence Erlbaum. Intentions 77 Snodgrass, J. G. (1989b). Sources of learning in the picture fragment completion task. In S. Lewandowsky, J. C. Dunn, & K. Kirsner (Eds), Implicit memory: Theoretical issues (pp. 259-282). Hillsdale, NJ: Lawrence Erlbaum. Spreen, O , & Strauss, E . (1991). A compendium of neuropsychological tests: Administration, norms, and commentary. New York: Oxford University Press. Squire, L . R. (1986). Mechanisms of memory. Science. 232. 1612-1619. Squire, L . R. (1992). Memory and the hippocampus: A synthesis from findings from rats, monkeys, and humans. Psychological Review, 99, 195-231. Squire, L . R., & Cohen, N . J. (1984). Preserved memory in retrograde amnesia: Sparing of a recently acquired skill. Neuropsychologia, 22, 145-157. Squire, L . R., Shimamura, A . P., & Graf, P. (1987). Strength and duration of priming effects.in normal participants and amnesic patients^ Neuropsychologia, 25. 195-210. Srinivas, K. , & Roediger, H . L . (1990). Testing the nature of two implicit tests: Dissociations between conceptually-driven and data driven processes Journal of Memory and Language, 29, 389-412. Tulving, E . (1972). Episodic and semantic memory. In E . Tulving & M . Donaldson (Eds), Organization of memory (pp. 382-404). New York: Academic Press. Tulving, E . (1983) Elements of episodic memory. Oxford, England: Clarendon Press. Tulving, E . (1984). Precis of Elements of episodic memory. The Behavioral and Brain Sciences. 7. 223-268. Tulving, E . (1985). How many memory systems are there? American Psychologist, 40, 385-398. Tulving, E . (1986). What kind of hypothesis is the distinction between episodic and semantic memory? Journal of Experimental Psychology: Learning, Memory, and Cognition. 12, 307-311. Tulving, E . , & Schacter, D. L . (1990). Priming and human memory systems. Science, 247. 301-306. Tulving, E . , Schacter, D. L . , & Stark, H . A . - (1982). Priming effects in word fragment completion are independent of recognition memory. Journal of Experimental Psychology: Learning, Memory, and Cognition, 8, 336-342. Intentions 78 Tulving, E . , & Thomson, D. M . (1973). Encoding specificity and retrieval processes in episodic memory. Psychological Review. 80. 352-373. Warrington, E . K , & Weiskrantz, L . (1968). New method of testing long-term memory with special reference to amnesic patients! Nature. 217. 972-974. Warrington, E . K . , & Weiskrantz, L . (1970). Amnesic syndrome: Consolidation or retrieval? Nature. 228. 629-630. Warrington, E . K . , & Weiskrantz, L : (1974). The effect of prior learning on subsequent retention in amnesic patients! Neuropsychologia. 12. 419-428. Watkins, M J , & Peynircioglu, Z. (1983) On the nature of word recall: Evidence for linguistic specificity. Journal of Verbal Learning and Verbal Behavior, 22. 385-394. Weiskrantz, L . (1985)! On issues and theories of the human amnesic syndrome. In N . M Weinberger, J. L . McGaugh, & G , Lynch (Eds), Memory systems of the brain (pp. 380-415). New York: Guilford. Weldon, M . S. (1993). The time course of perceptual and conceptual contributions to word fragment completion priming. Journal of Experimental Psychology: Learning. Memory, and Cognition. 19. 1010-1023. Weldon, M . S., & Jackson-Barrett, J. (1993). Why do pictures produce more priming on the word-fragment completion test? A study of encoding and retrieval factors. Memory and Cognition. 21. 519-528. Weldon, M . S., & Roediger, H . L (1987). Altering retrieval demands reverses the picture superiority effect. Memory and Cognition, 15, 269-280. Intentions . 79 Appendix A List of the sentences presented to subjects at study. Beside each cue is the sentence constructed using its respective target word. The sentences were constructed in such a way that no nouns or verbs occurred more than once. Cues Sentences Outdoor Sentences Word Associations Dogs: The alley cats clawed through the garbage. King: The boat in the harbor is named Queen. Wish: No stars were visible because it was so overcast. Working: It was hard to find our way through the woods. Black: It is unsafe to jog alone at night. High: There was a beautiful view at the top of the mountain. Loud: The passing tram made a lot of noise. Man: The women's rowing team placed first. Foot: While rock climbing I tore a hole in my new shoes. Moon: The setting sun was a blaze of color. Salt: I fell in the water when I went sailing. Soldier: The men in the trenches nearly died during the war. Stem Completions DR: We took a long drive through the country. RA: Walking in the rain is depressing. RO: There was a big bump on my side of the road. SA: She spent hours building a sand sculpture. BE: The bees were swarming around the bushes. CL: The gray clouds dumped a lot of snow on the skier. KI: We flew our kite at the beach. LA: The glacier lake was icy cold. BO: The boys checked their traps for rabbits. PL: He went to the park to play baseball. Intentions 80 S C : We went on a nature hike for our science trip. V I : The lawn grass was a vivid green. Indoor Sentences Word Associations Baby: The new doll can cry. Head: You should take some medication for that ache. Justice: The Christmas ornament spelled the word peace. Sit: The bowl fell down, out of the cupboard. Fingers: I keep my rings hidden in ajewel case. Gir l : I had my friends over for lunch. Needle: I need more thread to do my mending. Whiskey: He enjoyed a stiff drink after his exhausting day. Afraid: It is spooky when the basement is dark. Bath: The apartment was tiny with only one room. Chair: I placed the plates on the table. Sleep: Children like jumping on the bed. Stem Completions F O : When the freezer broke all the food went bad. J O : The family celebration brought great joy_ to all. Q U : When the newborn is sleeping we must be quiet. WI: My office does not have any windows. B L : I stacked my toy blocks all the way up to the ceiling. F A : Father spends too much time in his study. IN: Hide-and-seek is fun indoors. M O : My mother made me vacuum the carpet. H E : I use the help function on my home computer a lot. N E : He keeps his suite very neat. O R : The orange rug is so out of style. P A : Waxed paper is stored in the kitchen. Intentions 81 Appendix B NAME THAT WOR RESPONSE SHEET Subject #: Sex: M Education: List#: 1 SENTENCE RATING T A S K DIRECTIONS; This task is designed to familiarize you with the INDOOR/OUTDOOR category distinction. You will be shown 24 different sentences, one at a time. Please read every sentence and rate carefully (according to the scale below) the degree to which you think each sentence describes something or a situation belonging more to the INDOORS or to the OUTDOORS Rating Scale: 1 Completely indoors Somewhat indoors 3 Neutral Somewhat outdoors Completely outdoors 1 13 14 15 4 5-6 7 8 9 10 11 16 17 18 19 20 21 22 23 12 24 Intentions, 82 WORD A S S O C I A T I O N S A N D WORD S T E M C O M P L E T I O N S INSTRUCTIONS: Generate words for the following word association cues and word, stem cues according to the appropriate category label (eg., indoors, outdoors, or anything) provided. Please do not repeat any words. P R A C T I C E : INDOORS > Playing, INDOORS > D E OUTDOORS > Health OUTDOORS——• P O ANYTHING——• H A ANYTHING — • Rough S C O R E : . I M P E X P o INDOORS: - • : " '• INDOORS: _ _ » OUTDOORS: . ' ' ' .. • .,: . • OUTDOORS: .' • ANYTHING: ' • . • . ; • ® T O T A L o T O T A L Intentions « For each cue below, write down the first OUTDOOR word (i.e., a word associated primarily with the OUTDOORS) that comes to mind. Man • Loud • • C L . Needle . . B L . High . IN ; B E Whiskey F A ' K l . . • • ' . Girl . M O • L A ' Fingers Black Intentions 84 • For each cue below, write down ANY word (i.e., the first word that comes to mind; ANYTHING) that comes to mind. Salt •-. B O ' , Foot . Soldier 1 Afraid - ,. VI . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ O R • . ' •  SC N E . . . . , Moon • Chair • PL . Bath • H E •• -PA .Sleep . . . Intentions 85 • For each cue below, write down the first INDOOR word (i.e., a word associated primarily with the INDOORS) that comes to mind. Working DR , ; Q U . • Head • _ . Sit • R O WI • Dogs • S A ' . ' Baby . . ; Justice _ _ : F O • Wish , . K i n g ' JO . R A Intentions 86 R E C A L L T E S T INSTRUCTIONS: Think back to the sentences that you were asked to rate. For the following word association cues and word stem cues try to recall the emphasized words from the sentences that both fit the cue and that were presented under to the appropriate category label (e.g., indoors or outdoors) indicated Not all of the cues provided relate to the previously presented emphasized words, so think carefully. Please try not to repeat any of your responses. o For each cue below, write down an OUTDOOR word (i.e., a word associated primarily with the OUTDOORS) that you remember from the sentences previously presented and that was in a sentence associated with OUTDOORS. Man Loud C L • Needle B L High IN ' B E . Whiskey F A . KI Girl . M O L A . Fingers • Black Intentions 87 • For each cue below, write down an INDOOR word (i.e., a word associated primarily with the INDOORS) that you remember from the sentences previously presented and that was in a sentence associated with INDOORS. Working . D R Q U Head Sit . R O . • ' WI Dogs SA Baby Justice ' • F O Wish King : • JO .. R A ' • Intentions 88 Appendix C One of the major goals of this research laboratory is to find out about two different ways in which we use memory. These two ways have been called implicit and explicit memory. Explicit remembering occurs when we are consciously and intentionally trying to recollect information from a specific prior event or experience. An example of explicit memory use would be trying to remember what you had for dinner last night. In contrast, memory is said to be implicit when performance on a task is influenced by prior events or experiences in the absence of a conscious, intentional attempt to recollect those events or experiences. An example of implicit memory use would be when a tune "pops" into mind upon hearing the first few notes or when you are reminded of a particular person by smelling a perfume or cologne they frequently wear. Our research on implicit and explicit memory is motivated by several questions. The question addressed by this experiment is whether, and to what extent, these two forms of memory rely on the type of mental processing that occurs during a memory test situation. That is, do instructions to produce/recall words from specific categories or semantic sets influence implicit and/or explicit memory performance? For this purpose, we used word association and word stem completion tasks to test if responses differ according to the conscious constraints imposed by different types of instructions: (1) any word that comes to mind versus (2) only words related to things and activities associated with indoors versus (3) only words associated with things and activities occurring outdoors. The purpose of the sentence rating task was not only to focus attention on the indoor/outdoor orientation of the sentences, but also to focus attention on the emphasized words embedded in each of the sentences. Specifically, we wanted to see if any of these emphasized words would be reproduced in the word association and word stem completion tasks (measures of implicit memory) and, if so, under which category(ies) or semantic set(s) (i.e., anything, indoors, outdoors) specified by the retrieval instructions. Implicit memory performance will be compared to the conscious, intentional recall of the same words (a measure of explicit memory). The main independent variables in this experiment were the type of memory test (implicit or explicit), the type of task (word association task or word-stem completion task) and the type of instructions ~ whether performance on the tasks was guided by instructions to say any word that comes to mind or words relating only to indoors or outdoors. The main dependent variable was performance on the word association and word-stem completion tests, namely the number of words from the original study list that were produced both in the implicit and explicit test conditions. The critical issue of interest to us is the effect of the independent variables on performance on the memory tests. If you have any further questions, you may contact Angie Birt at 822-5121. If you are interested in the general topic examined by this research, you might read: Graf, P., Mandler, G., & Hayden, P. (1982). Simulating amnesic symptoms in normal subjects. Science. 8_ 415-431. Roediger, H. L. (1990). Implicit memory: Retention without remembering. American Psychologist, 45. 1043-1056. Again, thank-you for your participation. 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0087641/manifest

Comment

Related Items