Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

An experimental investigation of the impact of computer based decision aids on the process of preferential… Todd, Peter A. 1988

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

Item Metadata

Download

Media
831-UBC_1988_A1 T62.pdf [ 17.53MB ]
Metadata
JSON: 831-1.0098241.json
JSON-LD: 831-1.0098241-ld.json
RDF/XML (Pretty): 831-1.0098241-rdf.xml
RDF/JSON: 831-1.0098241-rdf.json
Turtle: 831-1.0098241-turtle.txt
N-Triples: 831-1.0098241-rdf-ntriples.txt
Original Record: 831-1.0098241-source.json
Full Text
831-1.0098241-fulltext.txt
Citation
831-1.0098241.ris

Full Text

AN EXPERIMENTAL INVESTIGATION OF THE IMPACT OF COMPUTER BASED DECISION AIDS ON THE PROCESS OF PREFERENTIAL CHOICE  by PETER A. B. COM.  TODD  1983, M c G i l l  University  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY  in THE FACULTY OF GRADUATE STUDIES Faculty  o f Commerce  and B u s i n e s s  Administration  We a c c e p t t h i s t h e s i s as conforming to the r e q u i r e d  standard  THE UNIVERSITY OF BRITISH COLUMBIA May 1988 0  P e t e r A.  Todd, 1988  In presenting degree  at  this  the  thesis  in partial  University of  freely available for reference copying  of  department publication  this or  thesis by  British Columbia,  of this thesis  or  requirements  I agree that the  for scholarly purposes  his  the  for  an  advanced  Library shall make it  and study. I further agree that permission for  her  representatives.  may be It  is  granted  by the  understood  that  head  extensive of my  copying  or  for financial gain shall not be allowed without my written  permission.  Department of The University of British Columbia 1956 Main Mall Vancouver, Canada V6T 1Y3  DE-6(3/81)  fulfilment of  Abstract This  research  examines  d e c i s i o n making p r o c e s s DSS  the  impact  of Decision  Support  Systems  f o r p r e f e r e n t i a l c h o i c e t a s k s . The  on the d e c i s i o n p r o c e s s  i s evaluated  i n terms of how  and  d e c i s i o n q u a l i t y . One  i n such a way o f the DSS  as  w i l l be  the system a l t e r s  three  w i l l be  on  used  o t h e r a s s e r t s t h a t the  geared towards e f f o r t c o n s e r v a t i o n .  s t u d i e s employed c o n c u r r e n t  the  developed based  assumes t h a t the DSS  on the d e c i s i o n process  the  r e s p e c t to d e c i s i o n e f f o r t  to maximise d e c i s i o n q u a l i t y . The  the impact o f the DSS The  l i n e of reasoning  on  p o t e n t i a l impact of  d e c i s i o n maker's c o g n i t i v e l o a d . Competing hypotheses are the p o s s i b l e o b j e c t i v e s o f the d e c i s i o n maker w i t h  (DSS)  use  These hypotheses about  are t e s t e d i n t h r e e  experiments. data  about  the d e c i s i o n p r o c e s s .  In experiment 1 s u b j e c t s were p l a c e d i n e i t h e r and  aided  or  setting  unaided  decision  alternatives strategy  from which  changed  subjects  as  behaved  a as  to  and  given  make a  results effort  of  2  was  similar  the  The  use  minimisers.  to  problems  choice.  r e l a t e d to the amount o f i n f o r m a t i o n Experiment  v e r b a l p r o t o c o l s to c a p t u r e  of  of  results the  There  either showed  five that  decision aid.  were  no  In  or  ten  decision general,  significant  effects  processing.  experiment  1  except  that  subjects  were  given  problems w i t h  e i t h e r t e n or  twenty a l t e r n a t i v e s . The  r e s u l t s were c o n s i s t e n t  with,  stronger  those  Almost  though  than  s u b j e c t s used E l i m i n a t i o n by Conjunctive  strategy.  minimisation. information  There  no  is  experiment  1.  strategy while  consistent  significant  with  a l l aided  group  the unaided group used a the  differences  notion in  the  of  effort  amount  of  processing  Experiment 3 was were a due  aspects  This were  of  to  the  designed  to t e s t whether the r e s u l t s  i n experiments 1 and  tendency o f d e c i s i o n makers to minimise e f f o r t  ii  2  or because  the  a i d was n o t p o w e r f u l  enough t o i n d u c e  the  DSS was a l t e r e d t o b o t h  increase  additive processing.  the support  f o r the additive  s t r a t e g y and reduce support f o r the e l i m i n a t i o n by a s p e c t s The  r e s u l t s o f experiment  3 show  d e c i s i o n makers  strategy  t o the type  additive  s t r a t e g i e s a r e made s u f f i c i e n t l y  employed.  Similarly,  elimination  strategy  way a s t o m i n i m i s e Overall the  when  the  aids  i s manipulated  a d a p t i v i t y centres  that  d e c i s i o n makers  make d e c i s i o n s faced  with  effort  and work  d e c i s i o n makers  sufficient  provided  changes  follow  tend  a  that i f be  particular  t o adapt  experiments are c o n s i s t e n t to the types  around  i n such  of support  the minimisation  are highly a way  by  conscious  i n such  a  i n demonstrating  tools available to  of decision effort. I t  o f the e f f o r t  as t o m i n i m i s e  the decision  i n the r e l a t i v e  d e c i s i o n makers  otherwise  their  t o use they w i l l  that  consider.  The  effort  a i d . There required  t o f o l l o w more e f f o r t f u l precise  nature  of this  required  expenditure.  the use a d e c i s i o n a i d they appear t o c a l i b r a t e t h e i r  to that  lead  to  t o adapt  effort.  a d a p t i v i t y o f d e c i s i o n makers  appears  effort  difference  i s evidence  less effortful of  study  approach. tend  a v a i l a b l e . There  degree  the r e s u l t s of the three  them. T h i s  can  of decision  that  In this  i s some  own  approaches  When  decision  evidence  t o use v a r i o u s than  effort-accuracy  to  that  strategies they  might  relationship  n e e d s t o b e s t u d i e d more c l o s e l y . The  basic  contribution  approach  f o r the  decision  theory  implications practical  study  and  of  the d i s s e r t a t i o n has  of  DSS,  information  f o r behavioural  implications  based  processing  decision  f o r the  on  settings.  iii  concepts  of  to provide  drawn  psychology.  theorists,  development  been  This  consumer  DSS  from work  a  formal  behavioural also  researchers  i n preferential  has and  choice  CHAPTER 1 - INTRODUCTION  1  CHAPTER 2 - THINKING IS HARD 2.0 I n t r o d u c t i o n 2.1 I n f o r m a t i o n p r o c e s s i n g and r a t i o n a l d e c i s i o n making 2.2 An overview o f e m p i r i c a l work on i n f o r m a t i o n l o a d CHAPTER 3- THE COST-BENEFIT FRAMEWORK 3.0 I n t r o d u c t i o n 3.1 The p e r c e p t u a l view 3.2 The c o g n i t i v e view 3.2.1 E f f o r t m i n i m i s a t i o n 3.2.2 A c c u r a c y m a x i m i s a t i o n 3.2.3 A c c u r a c y m a x i m i s a t i o n s u b j e c t t o an e f f o r t c o n s t r a i n t 3.2.4 E f f o r t m i n i m i s a t i o n s u b j e c t t o an accuracy c o n s t r a i n t 3.3 E f f o r t o r accuracy? 3.3.1 Conceptual models 3.3.2 S i m u l a t i o n models 3.3.3 E m p i r i c a l s t u d i e s 3.4 V a l i d i t y o f the c o s t - b e n e f i t framework 3.5 R e c o n c i l i n g the c o g n i t i v e and p e r c e p t u a l views  11 . 11 11 13 21 21 21 23 24 25 . . . . 26 . . . . 28 31 32 33 34 37 38  CHAPTER 4 - A BEHAVIOURAL DECISION THEORY APPROACH TO DSS DEVELOPMENT 4.0 I n t r o d u c t i o n 4.1 Models o f p r e f e r e n t i a l c h o i c e 4.1.1 The a d d i t i v e - c o m p e n s a t o r y model 4.1.1.1 Formal d e s c r i p t i o n o f s t r a t e g y 4.1.1.2 S u p p o r t i n g the s t r a t e g y 4.1.2 The a d d i t i v e - d i f f e r e n c e model 4.1.2.1 Formal d e s c r i p t i o n o f s t r a t e g y 4.1.2.2 S u p p o r t i n g the s t r a t e g y 4.1.3 The c o n j u n c t i v e model 4.1.3.1 D e s c r i p t i o n o f the s t r a t e g y 4.1.3.2 S u p p o r t i n g t h e s t r a t e g y 4.1.4 The E l i m i n a t i o n by aspects model 4.1.4.1 Formal d e s c r i p t i o n o f s t r a t e g y 4.1.4.2 S u p p o r t i n g the s t r a t e g y 4.2 Summary o f proposed support mechanisms 4.2.1 Computational support 4.2.2 Data storage 4.2.3 I n f o r m a t i o n d i s p l a y 4.3 Summary  . 41 41 44 46 47 50 51 51 53 54 55 56 58 59 60 61 62 69 70 72  CHAPTER 5- PROPOSITIONS AND HYPOTHESES 5.0 I n t r o d u c t i o n 5.1 P r o p o s i t i o n s 5.2 An overview o f dependent and independent v a r i a b l e s 5.3 H y p o t h e s i s development 5.3.1 E f f o r t M i n i m i s a t i o n Hypotheses 5.3.2 A c c u r a c y m a x i m i s a t i o n hypotheses  73 73 73 79 . 81 83 93  iv  CHAPTER 6- EXPERIMENTAL DESIGN & PROCEDURES 6.0 I n t r o d u c t i o n 6.1 E x p e r i m e n t a l d e s i g n 6.1.1 Experiment 1 6.1.2 Experiment 2 6.1.3 T e s t i n g f o r s c r e e n e f f e c t s 6 . 2 The t a s k environment 6.3 The s u b j e c t p o p u l a t i o n 6.4 The development o f t h e d e c i s i o n a i d 6.5 E x p e r i m e n t a l procedures 6.6 Data a n a l y s i s p l a n 6.6.1 B a s i c models 6.6.2 Dependent v a r i a b l e s 6.6.2.1 Computer l o g d a t a 6.6.2.2 V e r b a l p r o t o c o l d a t a 6.6.2.3 Q u e s t i o n n a i r e d a t a CHAPTER 7: RESULTS- EXPERIMENTS 1 and 2 7.0 I n t r o d u c t i o n 7.1 P r o t o c o l r e l i a b i l i t y 7.2 Experiment 1- r e s u l t s 7.2.1 E v a l u a t i o n o f assumptions o f s t a t i s t i c a l 7.2.2 P r e s e n t a t i o n o f f i n d i n g s 7.2.3 S a t i s f a c t i o n w i t h the system 7.2.4 Command usage 7.2.5 Problem s i z e 7.3 D i s c u s s i o n o f experiment 1 7.4 R e s u l t s experiment 2 7.4.1 Assumptions 7.4.2 P r e s e n t a t i o n o f f i n d i n g s 7.4.3 S a t i s f a c t i o n w i t h the system 7.4.4 Command usage 7.4.5 Problem s i z e r e s u l t s 7.5 Experiment 2- d i s c u s s i o n 7.6 Screen e f f e c t s - r e s u l t s 7.6.1 S t a t i s t i c a l t e s t s 7.7 Screen e f f e c t s - d i s c u s s i o n 7.8 P o o l e d r e s u l t s 7.8.1 H y p o t h e s i s t e s t s 7.8.2 Command usage 7.9 P o o l e d data- d i s c u s s i o n 7.10 C o n c l u d i n g comments  103 103 103 103 . 105 105 106 108 109 112 115 115 116 117 118 127  tests  CHAPTER 8- EXPERIMENT 3 : A 8.0 I n t r o d u c t i o n 8.1 Background and r a t i o n a l e f o r study 8.2 E x p e r i m e n t a l d e s i g n 8.2.1 High EBA, low AD ( c o n d i t i o n a l drop, no compare) 8.2.2 High EBA, h i g h AD ( c o n d i t i o n a l drop and compare) 8.2.3 Low EBA, low AD (no c o n d i t i o n a l drop, no compare) 8.2.4 Low EBA, h i g h AD (no c o n d i t i o n a l drop, compare) 8.3 Main e f f e c t s hypotheses v  130 130 130 133 134 135 142 145 148 148 156 158 159 163 164 167 168 171 172 175 175 176 180 181 187 189 189 191 194 198 . . 199 201 202 204  8.3.1 A d d i t i v e d i f f e r e n c e support hypotheses The compare f u n c t i o n The t a s k environment The s u b j e c t p o p u l a t i o n E x p e r i m e n t a l procedures Presentation of findings 8.8.1 The s t a t i s t i c a l model 8.8.2 A d d i t i v e d i f f e r e n c e r e s u l t s Command Usage Questionnaire r e s u l t s 8.8.4 E l i m i n a t i o n by a s p e c t s r e s u l t s Command usage Questionnaire r e s u l t s 8.9 D i s c u s s i o n 8.9.1 A d d i t i v e d i f f e r e n c e support Impact o f the AD support treatment Impact o f compare on memory L i m i t e d use o f compare Summary 8.9.2 EBA support S t r a t e g y changes Use o f c o n d i t i o n a l drop The exposure e f f e c t Summary 8.9.3 Combined e f f e c t o f EBA and AD support  8.4 8.5 8.6 8.7 8.8  205 214 218 219 . 219 221 222 223 227 229 232 235 236 236 239 239 244 246 248 249 249 250 251 254 255  8.10 Comparison o f r e s u l t s 8.11 I n t e g r a t i o n o f the f i n d i n g s 8.12 C o n c l u d i n g comments  259 263 271  CHAPTER 9 ANECDOTAL EVIDENCE 9.0 I n t r o d u c t i o n 9.1 A n e c d o t a l p r o t o c o l d a t a 9.1.2 Memory f u n c t i o n s and f o r g e t t i n g 9.1.2 I n f o r m a t i o n l o a d 9.1.3 A t t i t u d e s towards the system 9.1.4 Task d i f f i c u l t y 9.1.5 Summary o f a n e c d o t a l d a t a 9.2 A h y b r i d d e c i s i o n s t r a t e g y 9.3 C o n c l u s i o n s  272 272 273 274 279 284 289 292 293 298  CHAPTER 10- CONCLUSIONS 10.0 I n t r o d u c t i o n 10.1 Overview o f r e s u l t s 10.1.1 Process changes 10.1.2 Changes i n e f f o r t and i n f o r m a t i o n use 10.1.3 Impact on memory and a t t e n t i o n 10.1.4 A b e h a v i o u r a l d e c i s i o n theory i n t e r p r e t a t i o n 10.2 L i m i t a t i o n s o f the r e s e a r c h 10.2.1 I n c e n t i v e s 10.2.2 E x t e r n a l v a l i d i t y 10.2.3 S t a t i s t i c a l power vi  300 300 • 300 300 302 305 306 . 307 307 309 311  10.3 C o n t r i b u t i o n s o f the r e s e a r c h 10.3.1 C o n t r i b u t i o n s t o DSS r e s e a r c h 10.3.2 C o n t r i b u t i o n s t o b e h a v i o u r a l d e c i s i o n theory 10.3.3 P r a c t i c a l c o n t r i b u t i o n s 10.4 D i r e c t i o n s f o r f u t u r e r e s e a r c h 10.5 C o n c l u d i n g comments  3  1  2  3  1  3  315 3  1  6  3  1  7  *  3 2  TABLES  3  2  4  APPENDICIES  4  1  3  vii  L i s t o f Tables  T a b l e 4.1 A p p l i c a b i l i t y o f support f u n c t i o n s 325 T a b l e 5.1 P o s s i b l e impacts o f a i d on e f f i c i e n c y and e f f e c t i v e n e s s . . . . 328 T a b l e 5.2 Dependent and independent v a r i a b l e s 329 T a b l e 6.1 E x p e r i m e n t a l d e s i g n 332 T a b l e 6.3 Screen e f f e c t s d e s i g n 334 T a b l e 7.1 E x p e r i m e n t a l d e s i g n - experiment 1 336 T a b l e 7.2 T e s t f o r homogeneity o f v a r i a n c e ( E l * ) 337 T a b l e 7.3 P r o p o r t i o n o f i n f o r m a t i o n used ( v a l u e s a r e i n %) ( E l ) 338 T a b l e 7.4 A b s o l u t e i n f o r m a t i o n use ( E l ) . 339 T a b l e 7.5 P r o p o r t i o n o f a l t e r n a t i v e s examined i n d e t a i l ( v a l u e s i n p e r c e n t ) (El) 340 T a b l e 7.6 A b s o l u t e number o f a l t e r n a t i v e s examined i n d e t a i l (max. v a l u e s 5 and 10) ( E l ) 341 T a b l e 7.7 V a r i a n c e i n a t t r i b u t e usage ( E l ) 342 T a b l e 7.8 Mean number o f a t t r i b u t e s examined p e r a l t e r n a t i v e ( E l ) . . . . 343 T a b l e 7.9 Access p a t t e r n 344 T a b l e 7.10 S t r a t e g y assignments ( E l ) 345 T a b l e 7.11 T o t a l steps ( E l ) 346 T a b l e 7.12 T o t a l time 347 T a b l e 7.13 S a t i s f a c t i o n w i t h System 348 T a b l e 7.14 Ease o f use 349 T a b l e 7.15 Command u s e f u l n e s s ( E l ) 350 T a b l e 7.16 Command use 351 T a b l e 7.18 Problem s i z e e f f e c t 352 T a b l e 7.19 Design- experiment 2 353 T a b l e 7.20 T e s t s f o r homogeneity o f v a r i a n c e (E2*) 354 T a b l e 7.21 P r o p o r t i o n o f i n f o r m a t i o n used 355 T a b l e 7.22 A b s o l u t e amount o f i n f o r m a t i o n used 356 T a b l e 7.23 Number o f a l t e r n a t i v e s examined i n d e t a i l 357 T a b l e 7.24 V a r i a n c e i n a t t r i b u t e s examined p e r a l t e r n a t i v e 358 T a b l e 7.25 Mean a t t r i b u t e usage p e r a l t e r n a t i v e (E2) 359 T a b l e 7.26 Access p a t t e r n 360 T a b l e 7.27 S t r a t e g y assignments (E2) 361 T a b l e 7.28 T o t a l steps 362 T a b l e 7.29 T o t a l time 363 T a b l e 7.30 S a t i s f a c t i o n w i t h System (E2) 364 T a b l e 7.31 Command use (E2) 365 T a b l e 7.32 Problem s i z e e f f e c t 366 T a b l e 7.33 Amount o f i n f o r m a t i o n used 367 T a b l e 7.34 Number o f a l t e r n a t i v e s examined i n d e t a i l * 367 T a b l e 7.35 V a r i a n c e i n a t t r i b u t e s examined 367 T a b l e 7.36 Mean number o f a t t r i b u t e s examined 367 T a b l e 7.37 Access p a t t e r n 367 T a b l e 7.38 S t r a t e g y assignment 368 T a b l e 7.39 T o t a l steps 369 T a b l e 7.40 T o t a l time (seconds) 369 T a b l e 7.41 A b s o l u t e i n f o r m a t i o n use (pooled*) 370 T a b l e 7.42 Mean a t t r i b u t e s examined p e r a l t e r n a t i v e (pooled) 371 Table 7.43 Access p a t t e r n (-100 i n d i c a t e s pure attribute, +100 pure viii  a l t e r n a t i v e p r o c e s s i n g ) (pooled) 372 T a b l e 7.44 S t r a t e g y assignments 373 T a b l e 7.45 S t r a t e g y - a i d by s i z e 374 T a b l e 7.46 T o t a l time 375 T a b l e 7.47 T o t a l s t e p s 376 T a b l e 7.48 Command use 377 T a b l e 7.49 Experiment 1- summary o f r e s u l t s 378 T a b l e 7.50 Experiment 2- summary o f r e s u l t s 379 T a b l e 7.51 Pooled data- summary o f r e s u l t s 380 T a b l e 8.1 E x p e r i m e n t a l d e s i g n 382 T a b l e 8.2 T e s t s f o r homogeneity o f v a r i a n c e 384 T a b l e 8.3 I n f o r m a t i o n usage 385 T a b l e 8.4 A t t r i b u t e s a n a l y s e d i n d e t a i l 386 T a b l e 8.5 V a r i a n c e i n a t t r i b u t e s searched p e r a l t e r n a t i v e 387 T a b l e 8.6 Mean number o f a t t r i b u t e s searched p e r a l t e r n a t i v e 388 T a b l e 8.7 Access p a t t e r n (+100 t o -100) 389 T a b l e 8.8 S t r a t e g y - A d d i t i v e d i f f e r e n c e group 390 T a b l e 8.9 S t r a t e g y - EBA group 391 T a b l e 8.10 P r o p o r t i o n o f dependent e v a l u a t i o n s 392 T a b l e 8.11 P r o p o r t i o n o f dependent e v a l u a t i o n s - l a s t h a l f 393 T a b l e 8.12 T o t a l time (seconds) t o complete the t a s k . 394 T a b l e 8.13 T o t a l steps taken t o complete the t a s k 395 T a b l e 8.14 Command use- AD support treatment 396 T a b l e 8.15 R e l i a b i l i t y o f q u e s t i o n n a i r e c o n s t r u c t s 397 T a b l e 8.16 Q u e s t i o n n a i r e r e s u l t s - AD support group 398 T a b l e 8.17 Command use- EBA support treatment 399 T a b l e 8.18 Q u e s t i o n n a i r e r e s u l t s - EBA support group 400 T a b l e 8.19 Summary o f r e s u l t s - A d d i t i v e d i f f e r e n c e 401 T a b l e 8.20 Summary o f r e s u l t s : EBA support 402 T a b l e 8.21 Estimate o f s u b j e c t and system steps taken t o complete the task403  ix  A c k n o w l e dgement s  Many p e o p l e a s s i s t e d i n t h e c o m p l e t i o n o f t h i s w o r k . I n v a r i o u s ways t h e f a c u l t y a n d g r a d u a t e s t u d e n t s i n t h e b u s i n e s s s c h o o l h a v e h e l p e d t o s h a p e my i d e a s and s h a r p e n the f o c u s o f the r e s e a r c h . C h r i s Wagner was r e s p o n s i b l e f o r t h e c o d i n g o f t h e s y s t e m u s e d as t h e b a s i s f o r the s t u d i e s r e p o r t e d here. J a s b i r Singh helped w i t h data c o l l e c t i o n and c o d i n g o f the p r o t o c o l s . Each c h e e r f u l l y and k i n d l y t o o k t i m e f r o m t h e i r own p r o j e c t s t o a s s i s t w i t h m i n e . B a r b Weeks h e l p e d to t r a n s c r i b e the p r o t o c o l s , an o n e r o u s d u t y a t t h e b e s t o f t i m e s . N a n c y S h e l l made a v a l i a n t a t t e m p t a t i m p r o v i n g t h e r e a d a b i l i t y o f t h e f i n a l p r o d u c t . To a l l t h e s e p e o p l e a n d o t h e r s who a s s i s t e d i n v a r i o u s ways t h a n k y o u . My c o m m i t t e e members e a c h c o n t r i b u t e d i n v a r i o u s ways t o m a k i n g t h e d i s s e r t a t i o n b e t t e r a n d I h o p e t o m a k i n g me a b e t t e r r e s e a r c h e r . J o h n C l a x t o n a n d Don W h e r u n g b o t h f o r c e d me t o c h a l l e n g e my a a s s u m p t i o n s a n d question a p p r o a c h e s t o r e s e a r c h . T h i s h a s h e l p e d me t o b r o a d e n my p e r s p e c t i v e as a r e s e a r c h e r . A l D e x t e r p u t r e s e a r c h and d i s s e r t a t i o n s i n t o p e r s p e c t i v e , h e l p i n g me t o remember t h a t t h e r e i s a b i g g e r p i c t u r e t o c o n s i d e r . I z a k B e n b a s a t was my s u p e r v i s o r b u t much more t h a n t h a t he was a m e n t o r . I z a k r e p r e s e n t s a m o d e l o f how t o do r e s e a r c h t h a t a n y o n e c o u l d do w e l l t o e m u l a t e . H i s c o n t i n u a l i n v e s t m e n t o f t i m e a n d e f f o r t i n t o my d e v e l o p m e n t i s a d e b t t h a t w i l l n o t be e a s i l y r e p a i d . Thank you. F i n a l l y , C o n n i e , f o r s t a n d i n g b y me through the e n t i r e e f f o r t , your u n f a i l i n g s u p p o r t a n d b e l i e f i n me makes t h i s a l l w o r t h w h i l e .  x  CHAPTER 1 - INTRODUCTION  The  focus  of this  dissertation  support  systems  systems  a r e computer based  makers and a s s i s t no  strict  (DSS) on  i s on measuring  the p r o c e s s tools  the impact  o f d e c i s i o n making.  which augment  of decision  Decision  the c a p a b i l i t i e s  support  of decision  them i n s o l v i n g s e m i - s t r u c t u r e d d e c i s i o n problems. There i s  definition  o f what  constitutes  a  DSS;  however,  there  i s some  consensus on many o f the common c h a r a c t e r i s t i c s o f a DSS. They support  rather  than r e p l a c e d e c i s i o n makers, and focus on d e c i s i o n e f f e c t i v e n e s s r a t h e r efficiency  than  (Keen and S c o t t Morton, 1978). By the term s e m i - s t r u c t u r e d problems  we mean those problems f o r which t h e r e i s no known approach t o a r r i v i n g a t an optimal  solution.  The  normative  the  process  effectiveness  of  literature decision  on DSS emphasises making  in  order  i g n o r e d the process of decision making (Todd  and  Nault,  impact  1987). With the e x c e p t i o n  Morton,  to  enhance  and s u p p o r t i n g decision  making  (Keen and S c o t t Morton, 1978). Much o f the e m p i r i c a l work i n DSS  has  Scott  understanding  1971) most  research  o f a few i n i t i a l  i n the DSS  o f DSS on d e c i s i o n p r o c e s s e s .  and Benbasat, 1987; Benbasat  area  This research  studies  ( f o r example,  has n o t c o n s i d e r e d the focusses  closely  on t h i s  i s s u e . I t examines two q u e s t i o n s : 1) How can DSS be designed  t o support d e c i s i o n makers?  2) How does the use o f DSS impact The  first  the p r o c e s s o f d e c i s i o n making?  q u e s t i o n r e l a t e s t o system d e s i g n , the second i s one o f e v a l u a t i o n .  1  In  the  shortage (Hurt  approximately  of  et  empirical al.,  contradictory argued  that  perception Benbasat studies  of  h e l d by  these  history  the  DSS  and  there  has  been  evaluates  the  impact  the  same  time  the  and  Nault,  1987  for  contradictory  researchers  evaluating  of  examines  about the  (1987) have argued t h a t there  input-output The  At  Benbasat  many  year  work which  1986).  (see  20  evaluation an  results  due  DSS  literature  is  I t has  to  an  reasoning  b e h i n d the problem  ultimate  s o l u t i o n u s i n g a DSS.  o f DSS,  rather  outcome  between the  variables.  these  there  will  presentation  studies  typically  of  the  i s that i n r i c h ,  be  much  (Todd and  a  the  problem  Process  quality rather  d e c i s i o n problem  and  oriented for  studies  have  semi-structured  a  of  the  process  used  than  the  quality  of  by  the  a  capture  as a s e r i e s  better  priblems  chance  i t may  decision  outcome  its  of  a  of  well  be  maker  in  decision  Benbasat, 1987).  This  research  examines d e c i s i o n a i d s  alternative  p r e f e r e n t i a l choice  description  of  where a  cognitive  Input-output o r i e n t e d r e s e a r c h w i l l not  e f f e c t s . Also,  assess  evaluating  and  than the more common  these e f f e c t s , many o f which might negate each other when c a p t u r e d  to  Todd  i s a need f o r p r o c e s s t r a c i n g o r i e n t e d  f a i l u r e of input-output  contexts  occurring  easier  improper  p o t e n t i a l impact o f a DSS.  possible benefits  activity  capturing  been  studies.  ill-structured,  of  acute of  overview). are  an  the  which can  be  problems  c h a r a c t e r i s t i c s of  d e c i s i o n maker must described  by  1  choose  one  to  support m u l t i - a t t r i b u t e , m u l t i -  (Keeny  such of  and  Raiffa,  problems). a number o f  1976  These  are  provide  a  problems  a l t e r n a t i v e s each  of  a common s e t o f a t t r i b u t e s . Most consumer purchase  Throughout t h i s d i s s e r t a t i o n the term d e c i s i o n a i d or t o o l w i l l be i n t e r c h a n g e a b l y w i t h the term DSS. 1  2  used  d e c i s i o n s c a n be as  site  c h a r a c t e r i s e d i n t h i s manner. A l s o , b u s i n e s s  selection,  characterised answers  or  as  stock  purchase  preferential  choices  in  or  capital  choice problems.  that  the  choice  d e c i s i o n s such  investment  Typically is  problems  there  dependent  can  a r e no  upon  be  correct  individual  preferences. T h e r e a r e a number o f r e a s o n s f o r u s i n g t h i s p a r t i c u l a r c l a s s o f There  is a  l a r g e body o f  much o f w h i c h decision  focusses  makers  understood.  on  employ  This  determine  literature individual in  these  facilitates  appropriate  relating  ways  the of  d e c i s i o n behaviour. settings  providing  The  existing  e m p i r i c a l work  predict  how  decision  makers  these  types  Preferential with  by  choosing or  most  of  than  people  on  ignored  a  go  in  marketing  area  1976;  (Payne,  seems t o be  of  also  i s not  one  strategies that  documented  mechanisms  help  of  problems  us  to  decision  and  well  behaviour for  understand  aids  to  decision  designed  daily  a generic  basis  type  and to  (e.g.,  of problem which i s d e a l t  consumer  DSS  and  design  Painton  area  (likely  At  of and  DSS  but  Gentry,  that there  d e c i s i o n making;  the  because  same  behavioural  which i s r i p e  to say  purchase  i n a business  decisions, environment  f o r d i n n e r ) . I t i s a c l a s s o f p r o b l e m w h i c h has  i n the  f o r the  type  support  use  The  d e c i s i o n maker  between a l t e r n a t i v e p r o j e c t s or investments  implications  This  make  problems are  quantitative orientation).  problem  area  might  will  of  well  choice  problems.  choice  d e c i d i n g where t o  virtually  are  evaluation  makers.  support  to p r e f e r e n t i a l  problems.  of  time,  decision have not 1985).  i t s qualitative  studies  of  theory  have  pursued  Thus,  the  f o r the development o f a r e no  however,  those  3  this  research  rather type  do  exist  of  recognised i n the  preferential  DSS  choice  DSS.  systems which c u r r e n t l y support which  been  often  employ  this very  formal modelling  techniques.  o f many systems which  Humphreys and Wisudha (1987) p r o v i d e an  include formal  models  c h o i c e problems. These approaches may way  that  traditional  the  body  of  systems r e f e r r e d  some  normative  sense.  d e c i s i o n behaviour  solving  (Sprauge  chapter  4,  some  on  The  support  and  to as  these  They  tend  of  preferential  from DSS  i n the same  typically  separated  to  a  impose  d e c i s i o n maker which i s c o n s i d e r e d  DSS  approach  and  emphasises 1982).  systems  In  differs  in  flexible  that  experience  the  will  same  be  The  r e s e a r c h conducted  understand  based  understanding  decision  into  decision  making.  decision  makers  what  support  c o n s i d e r a b l e evidence  to  problem  discussed of  in  user  literature.  p a r t i c u l a r l y with  r e s p e c t t o the use  tools,  develop  objectives  Traditionally wish  to  i s a l s o r o o t e d i n the n o t i o n t h a t i n o r d e r to  d e c i s i o n maker b e h a v i o u r ,  computer  may  here  to  problems  r e s i s t a n c e t h a t are o f t e n r e p o r t e d i n the management s c i e n c e  "good"  i t attempts  approaches  a d d i t i o n , as  from  s t r u c t u r e or  the  Carlson,  of  DSS.  support  differentiated  management s c i e n c e models are  approach to problem s o l v i n g in  be  f o r the  overview  in  improve  we  must  d e c i s i o n makers DSS  the  research quality  of  hold  i t has  with been  their  i n the b e h a v i o u r a l d e c i s i o n t h e o r y  insight  and  respect assumed  to that  d e c i s i o n s . There literature  of  that  is this  not be  the case  (see f o r example, Payne 1982). T h i s r e s e a r c h examines the  assumption  implicit  i n most DSS  (1983) who  urged  research,  f o l l o w i n g on  the  advice  IS r e s e a r c h e r s to c h a l l e n g e the assumptions i m p l i c i t  of  Weick  i n their  work. For purposes o f t h i s cognitive and  l o a d . By  c o n s i d e r i n g DSS  then examining how  using  a  DSS,  r e s e a r c h DSS  i s viewed as a mechanism which  i n terms o f i t s impact  on  reduces  cognitive costs  d e c i s i o n makers might e v a l u a t e the c o s t and b e n e f i t s o f  i t i s p o s s i b l e to  make p r e d i c t i o n s o f  4  how  a  DSS  will  affect  d e c i s i o n s t r a t e g y . T h i s f o l l o w s the p e r s p e c t i v e , the  behavioural  trade  off  problem.  d e c i s i o n theory  effort  In  and  short,  accuracy  this  b o t h i n terms o f how  literature,  that  i n choosing  implies  that  the  c u r r e n t l y i n vogue i n much of  a  d e c i s i o n makers  strategy  for  a  selectively  given  d e c i s i o n maker e v a l u a t e s  decision  strategies  much e f f o r t i s r e q u i r e d to employ the g i v e n s t r a t e g y ,  what the b e n e f i t o f u s i n g t h a t s t r a t e g y may  be.  I n making such assessments the d e c i s i o n maker may  p l a c e d i f f e r e n t emphasis  (or weight) on e f f o r t , r e l a t i v e to a c c u r a c y . These d i f f e r e n t p e r s p e c t i v e s the d e c i s i o n maker may  adopt w i l l  to p e r f o r m c e r t a i n types o f a n a l y s i s . A  issues.  This  view  b e n e f i t s , was context  of  of  review o f  s e l e c t i o n i s provided  DSS,  which d i s c u s s e s  "marginal  economics  of  the  the  issue  of  discussed  effort."  In  that  3  development  cost perspective. supply  of  approaches limitations  of  in this  resources  decision  f a c i n g the  maker  development  strategies  of  strategies  are  is  decision  take  i n chapter  based  on  the  making  in  a  decomposed  Based upon these  that  such  to  a  sense  strategy  into a  notion  and  empirical  proposition. from a  cognitive  We  limited  restrict  summarise  "thinking  i s hard."  the these This  2. evaluation preferential  series of  decompositions  5  the  limitations  problem.  these  costs  d e c i s i o n makers have a s t r i c t l y  and  might  clarify  to  these i s s u e s i n the  i s a l s o viewed  d e c i s i o n maker i n the  i d e a i s developed more f u l l y DSS  research  I t i s assumed t h a t  cognitive a  DSS  to  required  relating  cognitive  work r e p o r t e d here can be viewed as a t e s t o f Keen's o r i g i n a l The  effort  literature  i n chapter  f i r s t advocated by Keen (1979) who  the  that  l e a d to d i f f e r e n t ways o f employing d e c i s i o n  s u p p o r t t o o l s , the p r i m a r y f u n c t i o n o f which i s to reduce the  d e c i s i o n maker s t r a t e g y  and  a  of  various  choice  behavioural  environment.  elementary  information  series of  decision  The  processes.  support  tools  are proposed f o r the v a r i o u s directly in  supports  the  s t r a t e g i e s . In t h i s way  generally  s o l v i n g p r e f e r e n t i a l choice  DSS  assumes  one  maker. T h i s  or  more  i s developed which  observed approaches taken by problems. Each o f the  elementary  i s done s i n c e  a DSS  cognitive  i t i s assumed t h a t  d e c i s i o n makers  functions  operations  from  c o s t s are reduced changes i n d e c i s i o n s t r a t e g y may  (Taylor,  approach  information initially  the  to  processing  DSS  by  decomposition  development  Newell  here draws h e a v i l y and  on  and  is  accuracy o f my  the  ideas  processes. (1979) and  the  decision  This  approach  o t h e r s who  DSS  When  rationality  (see  also  of  the  process to  assess  experiments tasks  system upon the  of  tracing the  contrast varying  the size.  of  behaviour Problems  6  the  system  human  that  was  development  Simon w i t h r e s p e c t  also  as  series  grounded  have advocated the  in  to of  more  decomposition  the r e l a t i o n s h i p between e f f o r t  Johnson and  Payne,  1985).  time t h a t such an  To  the  approach  DSS.  process  the  The  is  the f i r s t  techniques,  impact  in  strategies  developed i s employed i n a s e r i e s o f  use  protocols,  for  making  knowledge t h i s work r e p r e s e n t s  impact  studies  two  in decision  rooted  o f Newell and  of  been taken to the development o f a The  useful  processing.  d e c i s i o n behaviour  (1972).  o f d e c i s i o n s t r a t e g i e s as a means f o r s t u d y i n g  has  to be  the bounds o f  heavily  Simon  description  information  r e c e n t work by Johnson  best  decision  r e s u l t . In essence, systems  s e r v e to i n c r e a s e  approach to u n d e r s t a n d i n g  formalised  elementary  and  the  1975).  This  reported  into  the  i n o r d e r f o r a DSS  i t must reduce the c o g n i t i v e c o s t a s s o c i a t e d w i t h i n f o r m a t i o n  d e s i g n e d a l o n g t h i s p r i n c i p l e may  built  of in  three  d e c i s i o n making. The particular  decision  aids  of  and  of  aided 5,  experiments to  10  on  unaided and  empirical  concurrent strategy.  20  assess  verbal  The  decision  first makers  alternatives  are  considered.  Each  subject  solves  a p a r t m e n t s e l e c t i o n ) i n one without  the  use  concurrent strategy  of  verbal measures  the  of the  are  derived.  p r e f e r e n t i a l choice  aid.  While  that is  the  These measures  to  decision  hypotheses general  maker  are  d e c i s i o n and  holds  summed up  the  decision  i n the  m a n n e r i n w h i c h a DSS  may  (involving  the  information  other  usage  the  based upon the the  system.  assumption  amount o f e f f o r t assumption  as  concern.  a  primary  and  which  based upon the  quality  or  problem  hypotheses  d e c i s i o n makers i n u s i n g  f o l l o w i n g two be  solving  test specific  s e t s o f c o m p e t i n g h y p o t h e s e s : one  make a  are  which  d e c i s i o n maker i s p r i m a r i l y c o n c e r n e d w i t h  required  the  two  they  c o l l e c t e d from  were d e v e l o p e d about the b e h a v i o u r o f the There are  problem  three problem s i z e s e t t i n g s , e i t h e r with  decision  protocols are  a  that that These  propositions which r e f l e c t  the  used:  Pi  The use of the d e c i s i o n a i d to support the search process w i l l lead  the  d e c i s i o n maker to work more e f f e c t i v e l y ,  strategies  which  consider  information e v a l u a t i o n and  more  u s i n g more  information,  exhaustive  emphasise  show l e s s use of f i l t e r i n g  additive  and e l i m i n a t i o n  s t r a t e g i e s compared to unaided problem s o l v e r s .  P i ' The use of the d e c i s i o n a i d to support the search process w i l l lead decision  makers  to  work more e f f i c i e n t l y ,  using  the  least  effort  strategy which w i l l r e s u l t i n an acceptable s o l u t i o n .  The  r e s u l t s o f e x p e r i m e n t s 1 and  result with  of  use  effort  of over  the  DSS  and  accuracy.  2 show t h a t d e c i s i o n s t r a t e g y c h a n g e s as  that Such  these a  changes  result  7  is  appear  to  consistent  reflect with  a  much  the  concern of  the  l i t e r a t u r e developed  i n the b e h a v i o u r a l d e c i s i o n t h e o r y  Experiment 3 t e s t s and study of  a l l s u b j e c t s had  the  aid  provide  strategies. changes  It  access  the  another  the s u p p o r t e d  experiment  The  towards  show  that  research  to  advocated  research  addition, processes  of  has  by  Huber  (1982)  systems  problems  in  organisations  on  are a  credit  problems,  as  adaptive  to For  to adopt  environment they may The  adapt  the  to  results degree  known  a  focusses  for  is  consumer  implications i n that  specific  be  have a  of of  this  support  of  tasks.  the  choice  This  to  how  those  to  (Painton  rigourous  activities, notion  development  interest  about  on  i t provides  method  apply and  for  of  such  a as  has  been  Group  DSS.  i n the  consumer  computer  Gentry,  based  1985).  decomposing  In  decision  design.  numerous  daily  basis.  granting can  do  would  provides  as an i n p u t to DSS there  are  particular  and hence made e a s i e r to use  strategy?  which  than  to  work  of  f o r a g i v e n d e c i s i o n approach.  practical  rather  little  this  use  this  versions  the d e c i s i o n makers more i n c l i n e d  d e c i s i o n support  such  the  In  strategies.  where  and  for  d e c i s i o n makers  unsupported  area,  Finally,  location  on  previously  techniques  which  d e c i s i o n makers  structuring,  Development  support  provided  are  the  conducted  perspective  information  has  extent  supported,  p r o v i d e d f o r the v a r i o u s  choice  lesser  s t r a t e g y , even though i n an unaided  predisposition  new  or  s t r a t e g y i s more h i g h l y supported  i s not  E2.  to a d e c i s i o n a i d ; however, d i f f e r e n t  i n the degree o f support  example, i f one and  extends some o f the f i n d i n g s o f E l and  greater  tests  area.  For  any  can  preferential investment  a l l be  purchase  as w e l l as  8  of  example,  decisions  virtually  i m p l i c a t i o n s f o r marketing  examples  those  framed  choice  decisions, as  d e c i s i o n . Thus interested  made  in site  preferential this  i n DSS  research and  could  lead  to prescriptions  f o r the design  o f DSS w h i c h  wide range o f problems which c u r r e n t l y r e c e i v e The  dissertation  general that  literature  "thinking  process by  i s organised  which  i s hard."  Chapter  i s managed b y e m p l o y i n g a  looking  a t some  general  works  little  as f o l l o w s .  establishes  would be a p p l i c a b l e computer  Chapter  the v a l i d i t y  3  will  argue  cost-benefit which  make  2 will  This  assertion  e m p i r i c a l base found i n the p r e f e r e n t i a l choice  literature.  specific  make  evaluation  of the l i t e r a t u r e ,  about d e c i s i o n making o b j e c t i v e s and  accuracy  represent relevant  i n decision  Chapter  4  making.  will  by  discuss  focus  Together,  DSS  some  these  decision  two  making  making  will  b e done  and a l s o  a t the  Based upon general  on t h e d e s i r e  the development  behavioural  decomposing  elementary information various  will  assertion  this  arguments  to balance  effort  chapters  together  literature  which i s  research.  of actual  accomplished  of  which  an overview o f the behavioural to this  analysis  we  assess the  the decision  analysis.  this  support.  of the i n i t i a l that  to a  strategies decision  of  DSS  based  i n decision  strategies  upon  making.  into  a  This  their  careful will  be  constituent  processes. Given these d e c i s i o n processes the s t r u c t u r e  features,  to support  preferential  choice  problems,  c a n be  specified. Chapter 5 w i l l to p r e d i c t are  p r o p o s e two d e t a i l e d s e t s  a d e c i s i o n maker's b e h a v i o u r w h i l e  b a s e d upon t h e arguments  effort  minimisers  decision ideas  o f hypotheses which are intended  making  o f chapter  i n chapter  or accuracy  context  which  using  3 that  maximisers.  hypotheses  d e c i s i o n makers a c t e i t h e r as  They  i s established  t h e DSS. T h e s e  by  also  take  t h e DSS  into  account the  developed  from the  4.  Chapter 6 begins  the empirical  section o f the d i s s e r t a t i o n . I t presents,  9  i n d e t a i l , p l a n s f o r two designed  to  process  support  of  discussed  multi-attribute  decision in  experiments which t e s t the  making.  detail.  An  The  overall  preferential  dependent discussion  and of  impact o f the use choice  DSS  on  the  variables  are  decisions  independent  the  of a  experimental  paradigm  is  also provided. Chapter  7 presents  implications.  Results  the  are  results  of  these  discusses  a s s e s s e d and  two  their  experiments,  a pooling  o f the  data  analysed.  Chapter 8 p r e s e n t s an  a d d i t i o n a l experiment d e s i g n e d to answer a s p e c i f i c  q u e s t i o n based upon the major f i n d i n g s it  and  g i v e n i n d i v i d u a l l y f o r each o f the  the d i f f e r e n c e s between the experiments are is also  experiments  examines how  o f experiments 1 and  d e c i s i o n makers respond to d i f f e r e n c e s  2.  i n the  In p a r t i c u l a r , type o f  support  tools available. Chapter  9 provides  experiments. data  is  Evidence  provided.  s a t i s f a c t i o n and Chapter  10  contribution,  an  overview o f  which  supports  Additional  anecdotal the  findings  data  collected  interpretation relating  to  points  out  o v e r a l l summary o f  the  l i m i t a t i o n s of  areas o f e x t e n s i o n .  10  the  the  the  memory  a l t e r n a t i v e d e c i s i o n s t r a t e g i e s are a l s o  p r e s e n t s an  of  i n the  three  statistical load,  system  discussed.  d i s s e r t a t i o n assessing  work, and  discusses  its  possible  CHAPTER 2 - THINKING IS HARD  2.0 The  purpose  literature  of  this  showing  the  i n f o r m a t i o n l o a d and exhaustive,  but  come to  chapter  to  relationship  rather  the  is  is  cognitive  study  (i.e.,  Information  (1957)  that  Newell and  the  of  inducing  the  making  various  behaviour,  r e v i e w i s not meant to be wide  range  of  areas  and  Much o f t h i s d i v e r s e work  thinking, especially  first  p r o c e s s i n g and  formalized  behaviour.  He  c a p a c i t y of the  the  asserted  individual  activity.  with  It will  respect  to  a l s o show, to to  Taylor  (1975)  has  r a t i o n a l d e c i s i o n making concepts  of  bounded  that  to  the  due  i t i s impossible  maximizing)decisions  environments.  capacities  decision  of d e c i s i o n behaviour.  common c o n c l u s i o n  utility  processing  of  overview  strain.  "satisficing" processing  an  the ways i n which d e c i s i o n makers appear to adapt themselves  2.1 Simon  between  illustrative  problem s o l v i n g , i s a h i g h l y s t r a i n some e x t e n t ,  provide  o t h e r t a s k c h a r a c t e r i s t i c s . The  approaches used i n the has  Introduction  the  human  limited  in  all  but  the  argued  that  increasing  information  processing  most an  rational  simplistic individual's  d e c i s i o n making.  system  were  and  information  t o make t r u l y  c a p a c i t y would expand the bounds o f r a t i o n a l of  rationality  summarized  The by  Simon (1971) as:  -a r e l a t i v e l y slow s e r i a l p r o c e s s o r  ( m i l l i second speed)  -an e x t r e m e l y l i m i t e d s h o r t term memory (5-9 -virtually  infinite  long  term  11  memory  with  chunks) relatively  fast,  though  fallible,  access  on the o r d e r  o f m i l l i s e c o n d s , b u t slow  s t o r a g e (5-10  seconds f o r a u n i t o f i n f o r m a t i o n ) . These b a s i c making  limitations  behaviour  limitation (Miller,  t o c o n s t r a i n the problem  of individuals.  o f short  1956).  central  serve  reviewed  d e c i s i o n makers  these  note  i n particular  as a s e r i o u s p r o c e s s i n g below  a l l tends  are limited  must develop mechanisms f o r c o p i n g w i t h Since  researchers  term memory c a p a c i t y  The l i t e r a t u r e  conclusion:  Many  s o l v i n g and d e c i s i o n  bottleneck  to point  information  the  to  this  processors  who  information load.  p r o p o s i t i o n s have been p u t forward,  many d e t a i l e d e m p i r i c a l  s t u d i e s have documented p r e c i s e l y the manner i n which i n d i v i d u a l p r o c e s s i n g i s limited  or d e f i c i e n t .  F o r example,  people  v i o l a t e the fundamental axioms o f u t i l i t y a review). If not  have  theory  been  found  to c o n s i s t e n t l y  (see S l o v i c e t a l . (1977) f o r  These axioms d e f i n e the b a s i s f o r t r u l y r a t i o n a l d e c i s i o n making.  p e o p l e were t r u l y occur.  maximization  rational,  However, models  one o f the i m p l i c i t  i s that  T h i s i s c l e a r l y n o t the case. information environment  a t any g i v e n (Ackoff,  such v i o l a t i o n s  information  o f normative p r i n c i p l e s  assumptions  processing  would  of u t i l i t y  or value  i s a costless  activity.  D e c i s i o n makers a r e f a c e d w i t h an abundance o f  time.  This  1967; M i n t z b e r g ,  i s especially 1973).  true  Typically  in a  managerial  managers  have  no  problem g e t t i n g i n f o r m a t i o n b u t r a t h e r have a g r e a t d e a l o f d i f f i c u l t y  finding  relevant  limited  information  opportunity time. and  Given  t o focus  (Ackoff,  a t t e n t i o n on any p a r t i c u l a r  an environment  time p r e s s u r e  1967) . I n a d d i t i o n t h e r e  (Mintzberg,  that  is a strictly  problem  f o r any l e n g t h o f  i s c h a r a c t e r i z e d by b r e v i t y ,  fragmentation  1973) a d e c i s i o n maker would r e q u i r e  p r o c e s s i n g c a p a c i t y i n o r d e r t o make o p t i m a l d e c i s i o n s .  12  unlimited  2.2 Research  in  capacities  pool  or  the  has  (Kahneman, cognitive  An overview o f e m p i r i c a l work on i n f o r m a t i o n field  determined  1973) . effort  from  of  There  attention  that is  that  attention  some  question  processing  individual  is basically  available for distribution  specialized  studied  as  to  a  lower  than  that  required  for  the  resource amount  i s drawn from a g e n e r a l  modules  the  processing  limited  whether  (Allport,  1978;  However, the g e n e r a l c o n c l u s i o n t h a t the a b i l i t y to p r o c e s s level  load  typical  of  resource  Navon,  1985).  i s r e s t r i c t e d to a  d e c i s i o n problem  i s not  in  dispute. A  stream  increases  of  in  research  environmental  deterioration  i n the  by  Schroder  complexity,  amount o f  et  al.,  beyond  information  a  (1967)  demonstrated  threshold  processed.  limit,  Broadbent  led  w i t h more i n f o r m a t i o n they  restrict  salient  range  certain point maker such  to  cues  such f i l t e r i n g  focus  on  behaviour  processing.  of  important  will  Johnson  result  they  consider  can be  (1983) has  the  attention directing  information; in  to  the  however, a g a i n ,  omission  noted  that  of  the  problem.  causing beyond  relevant  behaviour  to  (1957)  n o t e d t h a t as d e c i s i o n makers are p r e s e n t e d the  that  has  tend Up  to  to  a  the  decision  a  threshold  information of  a  expert  from  decision  makers r e f l e c t s a tendency to overweigh i n d i v i d u a l cues which c o n t a i n abnormal or  outlying  data  values  Similarly,  Kahneman and  base  information  rate  heuristics  such  as  while  Tversky when  neglecting  (1982) note  making  availability  and  other  that  probability  t h a t which w i l l be emphasised i n t h i s r e s e a r c h . The Prospect  Theory w i l l  be  discussed  13  information.  i n d i v i d u a l s tend assessments,  representativeness.  i n showing such r e s u l t s have a somewhat d i f f e r e n t  their  relevant  to  instead  Kahneman and  ignore using Tversky  i n t e r p r e t a t i o n o f them than i n t e r p r e t a t i o n embodied i n  i n more d e t a i l  below  (Kahneman  and  T v e r s k y , 1979). In  the  mathematical  modelling  domain i t  i n d i v i d u a l s are v e r y  s e l e c t i v e i n t h e i r use  example,  in  a  problem  financial  prediction  about  the  company which  (Mears and  Firth,  1985).  i n general  (Slovic,  they have v e r y  1972). T h i s  systems a n a l y s t s .  has  processing Jacoby  area  of  (1974) was  research  a  a l s o been n o t e d  o f the  first  Given, f o r  with  2-4  i n t o t h e i r d e c i s i o n process (1986) w i t h  o f t e n be  respect  points  to  Increasing  cause r e l e v a n t i n f o r m a t i o n one  i n c r e a s i n g the  of  the  load  limited studies  information  i n marketing.  t h a t too much i n f o r m a t i o n the  to be  most  the  information  study  measurable when 10 performance due Such  a  alternative  to  interesting  findings  number o f a l t e r n a t i v e s l e a d s  (Malhotra,  considered 1982)  (Malhotra  indicated  that  et  of  performance A  to a t t r i b u t e s r e q u i r e s t h a t 15 or more be  requires  an  a  certain intuitive additional  iteration  14  appeal  since  through  an  this  to a d e c l i n e  a l . , 1982;  or more a l t e r n a t i v e s were p r o v i d e d .  f i n d i n g has  to  ignored.  be  attributes  had  l o a d leads  number  One  to  shown to outperform  p r o c e s s e d more r a p i d l y , than does an i n c r e a s e  1979).  well  experts  the amount o f i n f o r m a t i o n of  the  cues  domains as  v a r i a b i l i t y between  in Vitalari  to p o i n t out  consumers.  perspective  i s that  that  describing  estimated  seen i n o t h e r  individuals i s information  e f f e c t s on  DSS  noted  i s b e i n g m o d e l l e d (Dawes, 1972).  of  one  f i l t e r i n g which may From  reliably  self-insight  i n v e s t i g a t i o n which  capacity  dysfunctional  be  Simple mathematical models can  the d e c i s i o n maker who Another  information  i s considerable  little  been  dimensions an a n a l y s t w i l l make a  S i m i l a r r e s u l t s can be  (Dawes, 1979). In a d d i t i o n , there and  with  10-15  can  also  and w e i g h t i n g o f cues.  analysis  f i n a n c i a l p o s i t i o n o f a company a l o n g  has  in  i n the  Olshavky,  declines  were  s i m i l a r decline  in  presented. the  addition  individual's  of  an  choice  model  (assuming the d e c i s i o n maker does not  Evaluating  additional  attributes,  ignore  however,  the a l t e r n a t i v e e n t i r e l y ) .  requires  modifications  mental model used i n a n a l y s i s f o r e s t a b l i s h i n g weights and However, a d d i n g the  choice  provided  a  single attribute l i k e l y  model.  is  This  either  will  perceived  a t t r i b u t e s or to have l i t t l e fact  be to  results  especially be  relevance  i n only  true  highly  if  to use  the  relevance.  a minor change  the  correlated  to the t a s k .  t h a t , on average, d e c i s i o n makers tend  criteria  to  new  to  information  with  the  existing  T h i s i s r e i n f o r c e d by  the  only a small proportion of  the i n f o r m a t i o n cues a v a i l a b l e to them (Dawes, 1979). The for  the  fact  that a l t e r n a t i v e evaluation i s a binding  p o t e n t i a l of an  developing  Iterations  of  additional  s e t o f a l t e r n a t i v e s s h o u l d be  maker, whereas  automated  changes  to  or  d e c i s i o n aids  an  semi-automated  underlying  To  that  summarise  presented even  this  line  of research,  the  by  model  to  activities. deal  with  an  r e l a t i v e l y c o s t l e s s f o r the d e c i s i o n model,  which  supports  the  require considerable  decision  effort  and  can focus  decision aids, affect on  a  does appear to be  c h o i c e problems. I t has  such  consumer's  developing  there  a consensus  i n f o r m a t i o n s e t s c o n s i d e r a b l y s m a l l e r than  i n t y p i c a l r e a l world  simple  information, regard  these  individual.  t h a t consumers can be o v e r l o a d e d those  support  choice  maker's problem s o l v i n g s t r a t e g y , would s t i l l i n p u t by the  to  c o n s t r a i n t bodes w e l l  as  s o r t i n g the  choice  process  d e c i s i o n aids  to  a l s o been found  presentation (Russo,  support  of  1977).  preferential  price  In  this  choice  problems appears to be u s e f u l . D e c i s i o n maker responses to i n f o r m a t i o n l o a d can a d v e r s e l y a f f e c t d e c i s i o n quality  (Wright,  dysfunctional  1974).  behaviour,  Information such  as  overload errors  15  in  may  be  manifested  computations  or  either  in  comparisons  (Malhotra,  1984), o r  to  strain  reduce  reduction these  i n a reduced  (Jacoby,  1984).  i n decision quality.  would not  characteristics environment pressure decision  to  a l s o occur  by  make  making  a  final  information  of  to  be  processing  multi-alternative type  work.  is  likely  end  result  reason  themselves  be  clearly  In  in  i n consumer  the  information  a d d i t i o n much  pressure.  Given  a  that  d e c i s i o n environment.  richer  time  may  to s u s p e c t  are e s s e n t i a l l y  greater.  i n order  The  same.  The  and  the  managerial these  added  be f u r t h e r impeded.  d i s c u s s e d which  capacity  is  the  d e c i s i o n problems. Briefly,  the  is little  cases  i s o f t e n c h a r a c t e r i s e d by  area  case  managerial  i n both  decisions  p r e s s u r e s t h e i r a b i l i t i e s may The  i n the  manager  correct  either  which m a n i f e s t  o f the p r o c e s s o r  faced  In  Also, there  processing d i f f i c u l t i e s ,  problems,  quantity of information processed  i n three  A  shows study  study  experiments  evidence of  by  of  general  Payne  subjects  limitations  in  multi-attribute,  (1976) t y p i f i e s  this  were  with  presented  i n f o r m a t i o n about a number o f h y p o t h e t i c a l apartments d e s c r i b e d a l o n g a number of  attributes  the one The search  (such  as  rent, size  and  brightness)  apartment they would most p r e f e r f o r principal strategies  t a s k c o m p l e x i t y was  and  were asked  Payne study was  changed  task  a  function of  complexity.  to determine In  this  o p e r a t i o n a l i s e d as the s i z e o f the problem s p a c e  the number o f a t t r i b u t e s and a l t e r n a t i v e s  select  themselves.  r e s e a r c h o b j e c t i v e i n the as  to  2  how  case,  (i.e.,  f a c e d i n the c h o i c e problem).  Also  I t s h o u l d be noted t h a t i n t h i s case, and throughout the d i s s e r t a t i o n , the term problem space i s employed simply to r e f e r to the s i z e o f a problem facing the d e c i s i o n maker i n terms of i t s number of attributes and a l t e r n a t i v e s . T h i s i s a somewhat l o o s e i n t e r p r e t a t i o n o f N e w e l l and Simon's term "problem space," which i s used to i n d i c a t e an e n t i r e a r e a o f c o g n i t i v e space i n which a d e c i s i o n maker searches f o r a s o l u t i o n to a problem (Newell and Simon, 1972). C l e a r l y t h e r e i s a s t r o n g r e l a t i o n s h i p between the s e p a r a t e uses o f the term, but they s h o u l d not be viewed as s t r i c t l y e q u i v a l e n t . 2  16  of  interest  was  the  question  o f how  much  information  was  processed  (or  accessed) as a f u n c t i o n o f s i z e o f the problem space. The  findings  considered  declined  alternatives amount  indicated  of  both  sets,  alternative.  a  to  was  searched amount  study  design  use  percentage of  of  and  information  f o r each of  available  attributes  In addition, f o r small  The i n i t i a l  tend  the  the number  variable  to the e x p e r i m e n t a l makers  as  increased. information  alternative  that  information  the number sets  a  alternative.  information  was r e p l i c a t e d w i t h  was  constant  For  accessed  strategies  to  larger  f o r each  some minor m o d i f i c a t i o n s  and y i e l d e d e s s e n t i a l l y t h e same r e s u l t s .  additive  of  analyse  small  Decision  problems  and  e l i m i n a t i o n s t r a t e g i e s f o r l a r g e r problems. These r e s u l t s have a l s o been r e p l i c a t e d  i n other  t a s k domains.  (1979) extended Payne's work by examining v a r y i n g l e v e l s two d i f f e r e n t  types  of choice  technically  simple,  technically  complex,  clearly As  situation.  dichotomous  o f the problem  space  strategies  which  exploit virtually  strategies  which  emphasize  only  information searched in  for  search  the  with  Findings  study  subjects  move from  a l l available information i n the e a r l y  search  space.  appears  t o be  portion The  key  the number  "many reported  complexity. single-stage  to  multi-stage  and  a detailed  determinant  time  literature  cited previously.  Olshavsky a l s o  spaces the use o f s i m p l i f y i n g h e u r i s t i c s  per u n i t  of  information  17  examined.  of  of alternatives  r a t h e r than the number o f a t t r i b u t e s , a f i n d i n g c o n s i s t e n t w i t h  the consumer b e h a v i o u r  less  another,  attributes."  increases  filtering  reduced  i n this  that f o r l a r g e r search to  valued  and  i n d i c a t e t h a t the s t r a t e g y employed i s a f u n c t i o n o f t a s k  the s i z e  analysis  o f i n f o r m a t i o n under  One, which i s d e s c r i b a b l e by "many,  attributes"  multichotomous  Olshavsky  those found  generally l e d  In general,  this  study  confirms task  the  f i n d i n g s o f Payne (1976) and  extends them to problems o f v a r y i n g  complexity. Biggs e t a l . , (1985) examined the e f f e c t o f changing b o t h s t r u c t u r a l  variables  (e.g., t a s k s i z e ) and c o n t e x t u a l v a r i a b l e s (e.g., s i m i l a r i t y between  alternatives)  on  credit  granting  aspect  to  is  particularly realistic  the  decision strategy  decisions.  complexity.  processing  likely  germane  The  The to  i n that  contextual  more  be  adopted by  similar  required  for  with  consistent with  respect the  found t h a t a g r e a t e r  bank l o a n  manipulation two  to  problem  strategy problem  adopted space  proportion of  line  the  for  processed,  more f i l t e r i n g  still  complexity). and  relatively  work  small  (e.g.,  in  Biggs  settings,  is  made i n a  et  al.  (1985)  where  available,  d e c i s i o n makers would be  used  was  searched  as  requires  in  the  these less  greater  more  can  processing size  of  studies  the as  information  s t r a t e g i e s are  alternatives much  with  increases  adopted.  a is  It is  invoked when problems are by  four  attributes).  information  is  even more s e r i o u s l y c o n s t r a i n e d  o f e l i m i n a t i o n s t r a t e g i e s and  18  it  information  varies  often  size  (i.e.,  the  s t r a t e g i e s w i l l be  decision  were  common c o n c l u s i o n which  that  tasks  is  task  realistic  tendency towards the use  s m a l l . The  choice  six  is  thus has h i g h e x t e r n a l v a l i d i t y .  e l i m i n a t i o n type  to note t h a t these  more  T h i s i s c o n s i s t e n t w i t h the n o t i o n t h a t  (size As  new  the  This  a v a i l a b l e i n f o r m a t i o n was  research  considered  surrogate  a l s o important  of  in preferential  being  a  s t u d i e s c i t e d p r e v i o u s l y . I n terms o f s i m i l a r i t y ,  c o g n i t i v e e f f o r t ) when d i f f e r e n c e s are this  are  decisions being  d i f f e r e n t i a t i n g between o b j e c t s becomes more d i f f i c u l t  from  making  introduces  discrimination.  size  the a l t e r n a t i v e s became more s i m i l a r .  drawn  officers  alternatives  i t involves business  c o n t e x t by a c t u a l l o a n o f f i c e r s and  Results  be  task  In  typically and  the  simplifying heuristics  w o u l d be On  reenforced. the  whole  there  studies  conducted  the  that  fact  significant  is  a  central  i n a v a r i e t y of  "thinking  resources  complete  information  required  solution.  produce  makers e i t h e r by  a  design  or  that  and  emerges  with  This  the  s i t u a t i o n s which  enumeration,  and  from  individuals reduce the  is  i n d i c a t i o n that  from n e c e s s i t y  clear  are  highly  demand  resort  drastically a  many  f o r many d i f f e r e n t p u r p o s e s :  When f a c e d  strategies which f i l t e r to  domains  i s hard." for  conclusion  amount o f  cognizant  of  to work  decision  the  amount  o f e f f o r t t h e y expend on p r o b l e m s o l v i n g . The  findings  mechanism number  of  limits  and  dealt  which  causes  the in  cost  that  search.  general  effort required  i s how or  of  the  studies  these  raise  strain  have  been  reducing  put  detail  upon  are  the  there  here  is  the  made  marginal  i s thought  such  questions  as  The  the to  q u a l i t y more h i g h l y ?  particular interest  principle  assumption  d e c i s i o n makers are to increase  the  which  in  the  concerned with  place.  A  perceptual  processing  which w i l l  approach  which  return  to  from  a  be  trade  the  expected  will  deal with  other  words  is  the  much  decision  development  of  effectiveness  the and  DSS will  next  chapter  of  DSS  e f f e c t i v e n e s s o f t h e i r d e c i s i o n m a k i n g (Keen and  a  more  question since  literature use  effort  maker  This  of  made.  they value  or e f f i c i e n t d e c i s i o n s ? and  amount  choice  i n the  and  continued  o f f between the  a c c u r a c y of the  be  argues  strategies  design  19  the  problem-solving  t h a t we  underlies  take  of  about  t o be  In  nature  explanation  cost-benefit  t o implement a s t r a t e g y and principal  of  d e c i s i o n m a k e r s make t h i s c o s t b e n e f i t a s s e s s m e n t . Do  decision  of  issue  activities  forth  i n t e r e s t e d i n making e f f e c t i v e d e c i s i o n s is  the  t h i n k i n g (Payne, 1982).  decisions  based  In  of  some  procedures  One  these  propositions  with  generally  from  is  decision Scott  the that aid  Morton,  1978) . On  the  o t h e r hand many o f the c l a i m s r e l a t i n g t o the u t i l i t y  r e f e r more o f t e n t o t h e i r c o n t r i b u t i o n 1981).  I f we  decision that  consider  aiding  tools,  i t reduces  the  to e f f i c i e n c y i n d e c i s i o n making  spreadsheet,  one  i t s h o u l d become c l e a r  the c o m p u t a t i o n a l  load  of  of  the  that  most  commonly  i t s primary  DSS  (Keen,  employed  advantage  f a c i n g d e c i s i o n makers. I n t h i s  is  sense  such t o o l s are time s a v e r s . R e d u c t i o n i n c o m p u t a t i o n a l e f f o r t ,  resulting in a  time  further  savings  being invested  to  the  decision  i n problem  maker,  may  be  s o l v i n g . T h i s may  translated  into  then l e a d t o more e f f e c t i v e , w e l l  informed d e c i s i o n making. On the o t h e r hand these time s a v i n g s may other a c t i v i t i e s not  be  i n which case the d i r e c t b e n e f i t  apparent.  The  q u e s t i o n which must  effort,  be used f o r  from the d e c i s i o n a i d would  t h e n be  addressed  i s how  i s this  s a v i n g s i n time and e f f o r t u t i l i s e d by the d e c i s i o n maker? Does the d e c i s i o n maker have an o v e r a l l o r i e n t a t i o n towards i n c r e a s i n g quality of decisions order  to  minimise  p r e d i c t how  o r i s t h e r e a d e s i r e to s i m p l i f y the d e c i s i o n the  effort  associated  with  decision  making.  the  process i n In  order  to  a d e c i s i o n maker w i l l make use o f a computer based d e c i s i o n a i d we  must have some u n d e r s t a n d i n g o f the answers to the q u e s t i o n s r a i s e d above. In chapter these  3 empirical  evidence  i s reviewed which h e l p s us  issues.  20  to b e t t e r  understand  CHAPTER 3- THE  COST-BENEFIT FRAMEWORK  3.0 I n t r o d u c t i o n Normative of  utility  power as attempts man  as  models o f r a t i o n a l  d e c i s i o n making b e h a v i o u r , based  o r v a l u e maximization, have been found  descriptive to remodel  a limited  theories  of choice behaviour.  the concepts o f economic man  capacity  to have l e s s Recently  on n o t i o n s  than  adequate  t h e r e have been  t o f i t i n t o the framework o f  information processor.  I n g e n e r a l , the  thrust  of  t h i s work has been to develop more r o b u s t models which d e s c r i b e the b e h a v i o u r of  i n d i v i d u a l s making d e c i s i o n s .  broken the  into  two  other  a  approaches  cognitive  and  approach. We  groups:  focus  one  The  taking  approach.  in detail  i n these areas can be  a p e r c e p t u a l approach In  3  on  r e s e a r c h conducted  the  this  chapter  literature  t o d e c i s i o n making,  we  review  relating  to  a l s o review the evidence o f the r e l a t i v e importance  these  the  two  cognitive  of e f f o r t  and  a c c u r a c y i n d e c i s i o n making.  3.1 Prospect perceptual probably  theory  model  of  hardwired,  generalized  utility  The p e r c e p t u a l view  (Kahneman and  Tversky,  decision  behaviour.  approach  to  model.  Changes  the  1979)  I t argues  evaluation or  i s the  of  variance  that  best  example  there  is a  alternatives in  decision  of  a  single,  based  on  outcomes  a  are  a t t r i b u t a b l e t o the mechanism which i s used to e d i t the i n i t i a l r e p r e s e n t a t i o n of  the problem.  The  way  i n which the problem  Others (see Junngermann, 1985) p e s s i m i s t i c and o p t i m i s t i c approaches. 3  21  have  i s represented i s affected  referred  to  these  as  the  by  contextual variables gains  or  ( f o r example, whether  losses).  The  representation  in  These p r o c e s s e s  are thought  perceptual and  order  editing to  facilitate t o be  p r o c e s s i n g system.  way  which  prospect  not conform to normative the  decision  model.  For  process  serves  simplify  to  subsequent  systems  tend  theory  would  the  human  problem  evaluation  process.  e x p l a i n why  for  individual. decision behaviour being used to  implementation  perceptual  o f t h e human  to operate "automatically"  theory i s because the apparatus geared-up  of  the  the b a s i c l i m i t a t i o n s  conscious c o n t r o l o f the  i s not  example,  i s framed i n terms  the  g u i d e d by  These  are not under the d i r e c t , One  phase  the problem  system  of  attends  the  most  does  support  normative readily  to  changes o r d i f f e r e n c e s w h i l e the u t i l i t y or v a l u e models r e q u i r e p r o c e s s i n g i n terms  of  absolute  predictions that  o f the normative  the prospect  which  are  unaware  magnitudes.  theory  largely  of  and  differing  situations.  numerous  s t u d i e s where  shown t o  change  due  to framing If  theory  then very l i t t l e  If they  assistance w i l l  behaviour  may  control  Such  empirical  observations.  the  Individuals  strategies is  shown  large proportions  responses  differs  between To  the  i t would p r e d i c t d e c i s i o n  subjects.  invariance  very  be  is a  progress  making.  mechanisms w h i c h or  across  process  to questions,  they  by  of  would  use  decision  i n opposite  extent  behaviours  g e n e r a l l y be  (or  Kahneman  the  invoke)  and  Tversky  makers  have  directions,  for in been  simply  effects.  prospect  decision  result,  model i s v a l i d  to  their  a  model and  invariant  unable  As  complete  is likely  individuals use  and  manipulated  the in  model  of  decision  behaviour,  t o be made i n t e r m s o f a i d i n g o r  assisting  do  over  not  d u r i n g problem  facilitate  valid  have  conscious  control  solving,  t h e n no  amount o f  improvement this  22  case  by  of a  decision conscious  making. framing  the  training Decision of  the  problem to invoke c e r t a i n p r o c e s s e s from the d e c i s i o n maker, but  support t o o l s  would  not  be  virtually  prospect  theory  avenues maker  for  but  or  the  the  leads  may  Given  perceptual  exploration  rather  supported. This  worthless.  to  be  why  of  inference  we  view i n more d e t a i l . new  the  this  mechanisms  conclusion  for  that  will  consider  I t provides  supporting  decision  no  the  makers  some d e c i s i o n making r e s e a r c h e r s  useful  decision cannot  be  have r e f e r r e d to  t h i s as the p e s s i m i s t i c approach (Junngermann e t a l . , 1985).  3.2 An  The  c o g n i t i v e view  a l t e r n a t i v e view o f d e c i s i o n making i s p r o v i d e d  strategy  to be  particular  under  strategy  the is  conscious based  on  c o n t r o l of  the  some  of  form  by  those who  individual.  consider  The  "cost-benefit"  use  of  a  evaluation.  D e c i s i o n makers presumably c o n t r a s t the amount o f c o g n i t i v e e f f o r t r e q u i r e d  to  implement a p a r t i c u l a r s t r a t e g y w i t h the expected b e n e f i t s a s s o c i a t e d w i t h  the  particular  strategy.  measured as  the  The  benefits  l i k e l i h o o d of  accurate  response  decision  accuracy*  (Payne, a  that  1982).  trade-off  of  the  various  strategy Given is  leading  values  made.  strategies  The  are  typically  to a good d e c i s i o n or  for  cognitive  assumption  is  effort that  an and  ideally  d e c i s i o n makers would l i k e to maximise the q u a l i t y o f t h e i r d e c i s i o n s w h i l e a t the  same time m i n i m i s i n g  two  objectives  (Johnson  and  p r e c i s e l y how and  are  typically  Payne,  4  of  effort.  conflicting,  1985). The  However, to the some form o f  cost-benefit  effort  may  be  are  not  less,  than a  does  these  is  required  not  specify  Savings o f  single unit  The terms d e c i s i o n accuracy and d e c i s i o n q u a l i t y w i l l be used 23  that  t r a d e - o f f between e f f o r t  necessarily equivalent.  worth more, or  extent  trade-off  framework  d e c i s i o n makers view the n a t u r e o f the  accuracy. Trade-offs  unit  cognitive  a  single  sacrifice  in  interchangeably.  terms  of  evaluating  decision  quality  trade-offs  p r i m a r y emphasis  or,  3) are  between  a c c u r a c y . The  effort  and  accuracy  do  question  decision  is:  makers  when place  reduction?  increase  i n accuracy?  they i n d i f f e r e n t between the  Unfortunately, evidence w i l l be  central  on  1) e f f o r t 2)  or  the  literature  in  reviewed below. The  two?  this  area  t r a d e - o f f e v a l u a t i o n w i l l have i m p l i c a t i o n s  use  a decision aid. are  scant;  however,  existing  r e l a t i v e importance o f each component i n  the  Consider that there  is  f o r the way  four possible objectives  an  individual will  t h a t a d e c i s i o n maker  may  have: 1) minimise  effort  2) maximise a c c u r a c y 3) maximise accuracy s u b j e c t 4) minimise e f f o r t s u b j e c t These  latter  two  views  simply  t r y to  to an e f f o r t  to an a c c u r a c y express  the  constraint constraint.  f a c t that  a d e c i s i o n maker  may  place  r e l a t i v e l y more emphasis on a c c u r a c y or on e f f o r t . T h i s  we  shall  see,  has  possibilities particularly  will  Effort It  be  reviewed  interesting  b r i e f l y f o r the  3.2.1  implications  and  for but  decision only  defensible.  behaviour. the  The  Each  latter first  two  two  of  emphasis,  as  these  four  conditions  are  will  be  examined  sake o f completeness.  minimisation  is plausible  that  i n absence o f o t h e r  incentives  that  individuals  are  s i m p l y e f f o r t m i n i m i s e r s . They w i l l s e l e c t a d e c i s i o n s t r a t e g y which minimises  24  the  amount o f c o g n i t i v e  there  would  completing could  be  be a  criteria  task.  Thus,  defined  Completion  of  defined  the  the  no  as  a  e f f o r t required  as  imposed  for  a  the in  selection  requiring  the  quality  facing  a  from  the  generation  a possible  course  q u a l i t y or number o f a l t e r n a t i v e s put  forth.  In short, would be  generation of  on  example,  making  task  to complete a g i v e n t a s k .  the  idea  an  result  task,  choice  alternative  action  without  that  the  conjunctive  t h a t would be  (CONJ) or  set.  would regard  strategy  aspects  likely  strategies  be  or  evidenced  strategies  by  widespread  basing  choice  use on  of the  (EBA)  for  same  strategy,  approaches  extreme such an simple  random  examination  be  selection  an e f f o r t m i n i m i s i n g  e l i m i n a t i o n by  in  completion  undertaken w i t h o u t r e g a r d f o r d e c i s i o n q u a l i t y . Presumably, the  p r e f e r e n t i a l c h o i c e problems. However, taken to the would  end  potential  of e f f o r t minimisation implies  s t r a t e g y would always be used and such as  the  choice  of of  of  I n t h i s case  in  objective selection  of  a  single  criterion. The would  fact seem  1976). At accuracy  that  to  indicate  the that  empirical  very  that  least  must be  studies they  show  are  i t would  not  appear  m a i n t a i n e d and  that  that  decision  simply that  makers  are  e f f o r t minimisers  there  is  a  minimum  perhaps s t r a t e g i e s  are  adaptive (Payne, level  of  manipulated  to m a i n t a i n t h i s b a l a n c e (Bettman e t a l . , 1986).  3.2.2  Accuracy maximisation It  i s also  conceivable  that  decision  possible  d e c i s i o n i n a l l cases and  end.  fact,  In  makers w i s h  to  this  is  what  make o p t i m a l  a  makers would l i k e  to make the  best  t h a t they s e l e c t s t r a t e g i e s to a c h i e v e t h i s  normative  decisions.  25  decision  Violations  models of  assumes,  normative  decision  utility  and  p r e f e r e n c e models have been documented s u f f i c i e n t l y do  not  act  l i k e l y due  in a  utility  maximising  manner  maker  activity,  and as a r e s u l t i n d i v i d u a l s cannot  expect  must  of their to  see  operate.  solutions.  invariant  strategies being The  (Slovic  et  individuals  a l . , 1977).  This  is  i n p a r t to simple l i m i t s on i n f o r m a t i o n p r o c e s s i n g w i t h i n which the  decision  quality  to argue t h a t  Information  processing  not  a  costless  a t a l l times work t o maximise the  I f t h i s were the case  decision  is  behaviour  then,  with  once a g a i n , we  only very  complex  would  additive  used.  p r e c e d i n g two  l i n e s o f argument have g i v e n the extreme c a s e s . In  fact  s t r a t e g y s e l e c t i o n must be viewed as b e i n g somewhat more complex. I n d i v i d u a l s clearly  are a d a p t i v e  tend t o use for  i n the  strategies  they employ. For simple  problems  l i n e a r s t r a t e g i e s which most c l o s e l y approximate normative  l a r g e r problems s i m p l e r h e u r i s t i c s  are used which s a c r i f i c e  they  models;  some a c c u r a c y  i n d e c i s i o n making f o r reduced e f f o r t . C o n t i n g e n t d e c i s i o n b e h a v i o u r  i s one  of  the most s i g n i f i c a n t and r o b u s t f i n d i n g s i n b e h a v i o u r a l d e c i s i o n r e s e a r c h over the  last  ten  years  (Bettman  et  a l . , 1986).  o c c u r s c o n s i s t e n t l y a t t e s t s t o the f a c t  The  that  such  behaviour  t h a t d e c i s i o n makers are f o c u s s i n g on  t r a d e - o f f s between m u l t i p l e o b j e c t i v e s . I t i s argued to minimise  fact  e f f o r t and maximise a c c u r a c y  (Johnson  t h a t these o b j e c t i v e s are  and Payne, 1985). The  extent  to which e f f o r t and a c c u r a c y are v a l u e d i n t h i s t r a d e - o f f i s e x p l o r e d below.  3.2.3  Accuracy maximisation  One  line  quality  of  s u b j e c t to an e f f o r t  o f r e a s o n i n g would argue their  d e c i s i o n s but  information processing capacity. which  must  be  allocated  must  that individuals do  There  to v a r i o u s  26  constraint  so  within  the  is a  fixed  pool  activities  wish  to maximise  confines  of  the  limited  of c o g n i t i v e e f f o r t  (Kahneman,  1973).  It  seems  r e a s o n a b l e to a s s e r t t h a t e f f o r t w i l l be constrained  d e c i s i o n making. T h i s  argument. D e c i s i o n decision  within  makers  that  demonstrated  information  load  previous of  could  The  too  much  to see  d e c i s i o n makers use  was  reasoning  not would  i n the  in  a  larger,  employed. N e v e r t h e l e s s ,  considerably  some  imply t h a t  more  provided  cognitive  that  effectiveness  strategy  the  the  selection  was  otherwise  be  some left  consider  guided  of in  the the  that  the  by  to reduce c o g n i t i v e s t r a i n on  automating  causes studies  Olshavsky,  f o r simpler problem  1979;  problems and  as  a  would expect  a d d i t i o n a l expenditure Essentially this  a high  value  short,  the  on  of  line  of  obtaining  a  effort the  d e c i s i o n maker i s  impact o f a d e c i s i o n a i d i n s u p p o r t i n g  aid by  the  i s e s s e n t i a l l y bounded by  d e c i s i o n maker. In  First,  be  in  to expend a l a r g e amount o f  optimisation? should  to  oriented.  What would be where  the  d e c i s i o n makers p l a c e  or work put  c a p a c i t i e s of  1976;  complex  threshold.  i n t o a problem,  cognitive  due  o f the  under t h i s c o n d i t i o n we  to  Output,  the  reviewed  l o a d i s one  (Payne,  are w i l l i n g  so.  simply  that  m u l t i - a t t r i b u t e search  good, or c o r r e c t s o l u t i o n , and do  are  optimal  s t r a t e g i e s which are m o d e r a t e l y more e f f e c t i v e but  more e f f o r t ,  beyond  imply  literature  information  rationality  would  accuracy The  optimise  then make an  s t r a t e g i e s which work w e l l  effort  cannot be  effort  and  i n t e r p r e t e d i n t h i s manner  result  require  findings  and  argument  considerations.  a l . , 1986). C e r t a i n  require  space  This  effort  indicates that  behaviour.  c e r t a i n l y be  Biggs e t would  between  section clearly  contingent  s i m p l i f y a problem  overload  as to  i s e s s e n t i a l l y Simon's bounded  s i m p l i f i e d space.  trade-offs  or  expended i n such a way  computation hands  of  the  27  this  function  and  the  principle of  of  constrained  designed  decision  d e c i s i o n maker. I t does  storage  decision  a well  problem s o l v i n g  requirements  maker.  If  the  that  this would  individual's  objective  is  to  make  capacity,  in  an  aided  effort  i n t o the  to  best  too  decision  taxing  s e t t i n g s . Thus we  which more c l o s e l y  the  decision, decision  given maker  limited  would  f o r the  effectiveness.  selecting  expected v a l u e t h i s was  the  consistent who  conform w i t h  based  the use  normative models.  strategy  upon  equaprobable  more f r e q u e n t l y ,  approaches  o f more complex For  example,  rules  p r o v i d e d he  of  i s guided by  information  or  used i n  strategies  the  decision  outcomes as  would  likely  she  the  f o r a g i v e n problem. I n s h o r t ,  a c c u r a c y m a x i m i s a t i o n r u l e would use  implement more complex s t r a t e g i e s , c o n v e r t i n g  the  a h i g h e r q u a l i t y s o l u t i o n f o r the  g i v e n problem.  3.2.4  to an a c c u r a c y  E f f o r t minimisation subject Information  1980;  or  primary  well  use  understood  an that  " c o r r e c t " approach. T h i s would a l s o l e a d to the more complete  use  influences  geater  unaided d e c i s i o n maker would be  would expect to see  gambles  invest  Strategies  maker w i t h a d e c i s i o n a i d r e a d i l y combining p r o b a b i l i t i e s and as  processing  p o s s i b l y a c h i e v e a h i g h e r q u a l i t y r e s u l t . The  increase  which were s i m p l y  possible  environment  t a s k and  impact would be  aided  the  load,  strategy  Johnson &  or  overload,  is  s e l e c t i o n (Beach and  Payne,  1985;  Shugan,  clearly  In  a d e c i s i o n maker a decision  aid  expanded p o o l o f e f f o r t  to  into  constraint not  the  M i t c h e l l , 1978;  1980).  and  fact  only  factor  which  Christensen-Szlanski,  there  i s evidence  which  tends to i n d i c a t e t h a t s a v i n g s i n e f f o r t are more important to d e c i s i o n makers than  improving  decision  d e c i s i o n maker's t r u e constraint. sensitive the  to  In  this  the  q u a l i t y of  quality.  objective case  i t is  This  would  lead  to  i s to minimise e f f o r t argued  that  the  assertion  subject  decision  makers  to  an  are  that  accuracy much more  amount o f work which goes i n t o a problem than they are  the  resulting decisions.  28  Such t r a d e - o f f s  may,  of  a  course,  with also  be  affected  by  a variety  decision.  For  example,  important  than  the  maker would be the  decision.  of  other  the  decision  decision  to  maker would s t i l l be  the  to  rent  i n c l i n e d to p l a c e Within  conditions  an  of  as  purchase  a  apartment.  a higher  confines  such  level  the  importance  home  In  may  be  t h i s case  constraint  this constraint,  on  expected to be more s e n s i t i v e to i n c r e a s e s  whole i s s u e  most  easily  of  be  determines  the  level  interpreted  the  initial  of  incentive  as  an  input  constraint  on  p r o v i d e d to into  the  decision  the  decision  maker  will  then  be  s e l e c t i n g an a c t u a l s t r a t e g y This might be  implies the  that  freed  into  decisions.  with  smallest  automate a  series  work  the  on  From  effort  between m o d i f y i n g  the  problem  a decision  strategy  a i d , we  reducing  effort. the  must a l t e r the example,  a  quality  the  considerations  in  of  variable,  the  than c o n v e r t i n g  the  may  an  If a  the  makers  as  load,  lead  to  acceptable decision  better the  solution  a i d were  effort for  each  to but  would a n t i c i p a t e  same s t r a t e g y  as used i n  d e c i s i o n a i d to induce change i n t h i s case, i t  e f f o r t r a n k i n g s or r e l a t i o n s h i p between v a r i o u s  i f decision  maker  would expect t h a t  cognitive  d e c i s i o n maker would c o n t i n u e to u t i l i s e  an u n a i d e d environment. For  the  may  of  m a i n t a i n i n g t h e i r r e l a t i v e degree o f d i f f i c u l t y or s t r a i n , we t h a t the  l e v e l . Thus,  set  either  which  which p r o v i d e d  expenditure of  strategies,  decision  decision  level  f o r the l e a s t e f f o r t r o u t e r a t h e r  possible of  to  the  of  i n the problem s o l v i n g p r o c e s s .  choice  Thus, when u s i n g  maker would use  sensitive  quality  a d e c i s i o n a i d which reduces c o g n i t i v e  extra  decision the  a  case when u s i n g  resources  quality  f o r use  given  d e c i s i o n maker would opt  more  the  quality.  s t r a t e g i e s which meets t h i s minimum a c c e p t a b l e  decision  d e c i s i o n maker  way  decision  the  more  i n e f f o r t than  to adjustments i n d e c i s i o n q u a l i t y beyond the minimal c o n s t r a i n e d the  the  much  the  however,  of  were  i n d i f f e r e n t between  29  the  strategies.  application  of  For a  conjunctive  or  alternatives,  elimination  by  aspects  a d e c i s i o n a i d which made the  might r e s u l t i n a s h i f t which e q u a l l y  i n strategy.  Shifting  of  required  e f f o r t can be assumed by required  strategies to use  the  aid  the  cost  i s e s s e n t i a l l y the  would  and  i n making  a  choice  approach e a s i e r  to  i n a decision aided  changing the  effort  between implement  environment  relationships  strategy.  require  that be  net  difference net  the  the  number  aid.  system. The  The  net  of  in  difference  t o t a l number o f c o g n i t i v e  using  the  the  p o s i t i v e . The  subtracting  function  of using  units  latter  the e f f o r t r e q u i r e d  that  the  two  to implement v a r i o u s  Provide  degree  out  tools  with  ways to  of  value  can  to which they assume c o g n i t i v e  be  d i f f e r e n c e proposed here  create  levels  in  effort  effort"  differences  between  s t r a t e g i e s w i t h a d e c i s i o n a i d are  differing  the  operations  same as Keen's n o t i o n o f "the m a r g i n a l economics o f  1979) . I t p o i n t s  1)  shift  d e t e r m i n i n g the  decision  implement  thought o f as  (Keen,  In s h o r t ,  a c e r t a i n strategy  a s s e s s e d by  the  to  EBA  s u p p o r t e d a l l s t r a t e g i e s , not  between them, there would be no  effort  strategy  of  power  operations  i n terms f o r the  of  to the  decision  maker or, 2)  Provide  interface  techniques  i n h e r e n t l y more d i f f i c u l t Of  course,  approach  in  2),  necessary.  however  Consider,  maker  to  task.  I t may  make  virtually  use  an not  i t easier  unsupported.  In  f o r example,  additive be to  are  strategy  possible use  than  these  cases  approach  situations the  over  an  elimination  i t may  be  30  to  certain  1)  of  would  are  f o r the  approach,  be  preferable  to  approach  2)  t r y i n g to get  elimination  the  functions  others.  where  difficulty  to develop a i d s an  which  to employ than  a l l problems  there  in  strategy  additive even  would a  decision  in a  strategy  i f the  system d e s i g n e r ' s  be  choice which  latter  was  advantage  to  a c t u a l l y put helping  to  impediments i n the way t i p the  effort  This  argument p o i n t s  must  consider  provide  and  the  that  differential  to the  both  f a c t that  costs  this  summarise,  reasoning  that  there  accuracy  literature  r e l a t i n g to  difference  three  makers  are  optimisers. effort  A  indicate making  that  q u a l i t y . The 1)  three  research  heuristics  3)  highly  those  tools  important.  of  logical  are  we  likely  to  will  be  s u p p o r t both  the  This  argument to  minimisers  issue  and  examination  trade-offs  will  that  of  help  they  are  of  the  some to  shed  further  cost  accuracy? which we  can  examine to h e l p  untangle the  however, taken t o g e t h e r i t does tend to conscious  relatively  models  approach.  less  of  the  effort  concerned w i t h  w i l l examine benefit  they  optimising  put  into  decision  are  trade-offs  (Beach  and  Mitchell,  Shugan, 1980);  simulation  1986);  are  additive  a c c u r a c y to d e c i s i o n makers. None o f  conclusive;  areas o f work we  conceptual  1978; 2)  and  the  i s correct.  E f f o r t or  i n d i v i d u a l s are  decisions  is  closer  accuracy  l i n e s of  considered  that  effort  the r e l a t i v e importance o f e f f o r t and e v i d e n c e can be  of  4.  3.3 are  i n favour  benefits  l i g h t on which o f these a s s e r t i o n s  There  o f e l i m i n a t i o n s t r a t e g i e s , thus  i n developing d e c i s i o n a i d i n g tools  i s evidence and  decision  constrained  and  net  developed f u r t h e r i n c h a p t e r To  o f the use  work  on  (Thorngate,  the 1980;  effort  and  Johnson and  accuracy Payne,  of 1985;  various  decision  Bettman e t  al. ,  and  empirical  (Russo and  studies  examining  Dosher, 1983;  strategy  Bettman and  31  selection  Kakkar, 1976;  in  choice  Jarvenpaa,  problems 1988)  3.3.1  Conceptual models Shugan  accounts see  (1980)  develops  f o r expenditures  significant  use  a  conceptual  of  cognitive  of u t i l i t y  f a c t t h a t those s t r a t e g i e s f a i l to  strategy  attempting decision.  s e l e c t i o n . He to minimise  He  confidence  expresses  level  this  effort.  or v a l u e  that  subject  a  to  correct  He  the  in  constraint  f o r making  of  strategy  the  strategies  failure  i s due  to  " c o s t o f t h i n k i n g " as an  selecting  strategies  a constraint in  s e l e c t i o n which  argues t h a t  maximising  to c o n s i d e r  asserts  effort  model  terms  of  decision.  on  an  the  to the  input  people  are  q u a l i t y of  the  externally specified  Strategies  are  evaluated  in  terms o f the amount o f e f f o r t r e q u i r e d to meet the s p e c i f i e d l e v e l o f d e c i s i o n quality.  The  w i l l be  least effortful  strategy  which p r o v i d e s  chosen. Shugan (1980) a l s o r e c o g n i s e s  an  acceptable  t h a t e f f o r t may  be  task  c h a r a c t e r i s t i c s such as  s i m i l a r i t y between a l t e r n a t i v e s or  the  decision.  the  individuals o f the  In  general,  model  a c t to minimise e f f o r t  is  subject  consistent  with  specify  the  Beach  and  Mitchell  contingencies  (1978) have  which  may  developed  effect  the  a  the  s e l e c t i o n . For  task  will  influence  strategy  such as c o m p l e x i t y , n o v e l t y , desire  to make an  problem.  maker may the  the  At  idea the  of  that  quality  designed of  decision  d e c i s i o n maker  example,  task  to  and  conditions  ambiguity, or s i g n i f i c a n c e o f outcomes can e f f e c t  accurate  the  model  selection  c h a r a c t e r i s t i c s o f the  the  importance  to some c o n s t r a i n t on  s t r a t e g i e s . They argue t h a t v a r i o u s  on  a f u n c t i o n of  decision.  Similarly  the  decision  same  i n f l u e n c e h i s or her  choice time  the  or  the  amount o f e f f o r t  general  capabilities  t o be  of  d e s i r e to work on a problem and may  type o f s o l u t i o n t h a t w i l l be  accepted. I t should  32  be  the  focussed decision  also  noted that  impact  cognitive  load  is  but  one  aspect  of  the  model.  Beach  and  Mitchell  d e c i s i o n maker i s m o t i v a t e d to s e l e c t the  least effortful  provide  various  an  acceptable  involves  the n o t i o n  and  conflicting  the  s o l u t i o n given  the  Simulation Thorngate  examined  desire  to  minimise  (1980) conducted a  the  q u a l i t y of  simulation  a t the this  the  effort  simulation  expended  though not  tendency  to  optimal  on  ignore where  the  or  relative  to  by the  clearly  on  the  problem.  The  the  the use  decision  of  in  considered  fact  that  to  an  the  usefulness  e x p e c t e d model.  information  the and  than the  load.  The  do  so  maker.  He  while  simply  effort  justified.  The  optimal  h e u r i s t i c s while,  at  imposing  argues  cognitive  the  expected value  heuristics efficient.  p r o b a b i l i t i e s may  increase  to e v a l u a t e  The  results  the h e u r i s t i c s performed r e l a t i v e l y w e l l  solutions  misuse  s o l u t i o n i s not  motivated  This  to an a c c u r a c y c o n s t r a i n t .  relative  l e v e l s of  (1980) c o n s i d e r s  optimal,  demands  trade-off,  be  study  same time r e q u i r e d much l e s s e f f o r t  cognitive  will  employed to make d e c i s i o n s under r i s k .  decisions  indicated that  sense Thorngate  good,  s t r a t e g y which  models  h e u r i s t i c s were t e s t e d under v a r i o u s o f the  the  o f a t r a d e - o f f between the d e s i r e to make a good d e c i s i o n  o f a v a r i e t y o f h e u r i s t i c s which can be He  that  contingencies.  i m p l i e d d e c i s i o n r u l e i s to minimise e f f o r t s u b j e c t  3.3.2  state  use  model the  to  model. In  They  provide  relatively  that  an  be  the  move  and  few  individual's result  from  a  good  to  o f the h e u r i s t i c s  may  provides  same time  a  of  little imposing  benefit a  large  cost. Johnson and extended decisions  the  Payne (1985) conducted a s i m u l a t i o n study which r e p l i c a t e d and  work o f Thorngate. While  produced by  the  the  latter  h e u r i s t i c s , Johnson and  33  focussed  on  the  q u a l i t y of  Payne modeled b o t h  effort  and  accuracy  trade-off  of  between  heuristics  over  alternatives, probability the  choice The  the  number  of  the  that  of  involved  However,  Payne argue t h a t  the  contingent review).  This  the  the  nature  performance  conditions  by  for  alternative,  each  absence o f  suggest  using  Also,  varying  the  of  the  of  the  number  of  variance  in  dominated a l t e r n a t i v e s  that  some  there  heuristics  information  makers  the  need  single can  to  load be  Heuristics  in  little  point  e f f o r t while  The  empirical  at  of  in  the  case.  studies  34  The  in  findings  order  to  use a  well  under  all  while  saving  the  (see  that  i n terms o f  evidence to be  stable  normative approach. Johnson  notion  same  of  same time m a i n t a i n i n g  at l e a s t a p a r t i a l  view  trade-off  accuracy  increases.  i n many s t u d i e s the  the  explanation  for  Payne, 1982  for  decision  have been developed which g r e a t l y  maker's  t h i s i s i n l a r g e measure the  Empirical  reinforces  sacrificing  decision minimise  found  a  relatively  accurate  e f f o r t r e l a t i v e to the  also  and  performs  highly  are  indeed  adaptive  heuristic be  these t r a d e - o f f s  same time  the to  No  is  some h e u r i s t i c s are  as  heuristics  work  l e v e l s of accuracy.  3.3.3  in  d e c i s i o n behaviour  conscious.  heuristic  outcomes  decision  accuracy.  and  from  understand  simulated  task  simulation  d e c i s i o n maker s u b s t a n t i a l  key  to  to e f f e c t i v e l y reduce e f f o r t w h i l e a t the  conditions.  at  of  e f f o r t they r e q u i r e  degree  while  also  presence or  they p r o v i d e .  indicate  effort  attempt  They  of  the  the  effort  that  heuristics  a  an  set.  responses  high  two.  l e v e l s , and  the  terms o f  the  in  a wide v a r i e t y  results  between  also  heuristics  is  to  time  are  reduce e f f o r t  decision chose  makers  quality.  the  maintaining  The  appropriate reasonable  reviewed below suggests  that  Russo and Dosher (1983) i n t h e i r study o f b i n a r y c h o i c e problems have a l s o asserted  t h a t i n d i v i d u a l s make d e l i b e r a t e c h o i c e s  upon a t r a d e - o f f between e r r o r  and e f f o r t .  show  effort  a  greater  maximisation. Individual  attention  to  Christensen-Szalanski  d e c i s i o n making  They c l a i m  reduction (1980)  i s adaptive  information  load  considerations  amount o f e f f o r t p u t i n t o a problem al. of  (1986) demonstrate t h a t strategies  perform  as  decision  information decision  that  (1977)  incentives  arguments.  1980). Bettman e t  and  a relatively  show  they  constant  expend  allow  People  level  would  These  change  findings  (1972) c l a i m  that  on  strategies  i f accuracy  information  fell  presentation  indicate  a  that  (1987) has a l s o them t o p r o c e s s  below  a  that  certain  are consistent  d e c i s i o n makers tended t o use i n f o r m a t i o n  notion,  of  i n finding  r e s u l t s which  that  i n the types  accuracy.  important. J a r v e n p a a  d e c i s i o n makers use s t r a t e g i e s  c a n impact the  adaptive  effort  effort  also  e f f o r t minimisation  form p r e s e n t e d . T h i s  belief  the  accuracy  i n the l e a s t e f f o r t f u l manner. A t the same time i t was n o t e d  makers  threshold.  to maintain  minimising  and Kakkar  consider  demonstrated  the  while  Bettman  individuals  are t r y i n g  similar  (Christensen-Szalanski,  both  to  t r a d e - o f f s a r e n o t simply due  external  considering  do  trade-offs  a t r a d e - o f f between the  d e c i s i o n makers a r e h i g h l y  select,  i f they  quality  solution.  Slovic's  they  these  they  makes  and r e f l e c t s  since  that  than  also  c o s t s and b e n e f i t s o f making a d e c i s i o n . F u r t h e r to  o f d e c i s i o n s t r a t e g i e s based  with  only i n  l a b e l e d c o n c r e t e n e s s , was a l s o based on the  t h a t d e c i s i o n makers a r e h i g h l y c o n s c i o u s o f the degree o f e f f o r t  that  they p u t i n t o a problem. To being  summarise conclusive,  reasonable  levels  these  arguments,  there  i s evidence,  though  i t i s f a r from  t h a t i n d i v i d u a l s s o l v e problems i n such a way as t o m a i n t a i n o f accuracy  while  35  a t the same  time  trying  t o minimise  effort.  In  this  repertoire  of  case  we  expect  strategies  in  order  o p t i m a l s o l u t i o n s w h i l e a t the a  s i m i l a r review  primary o b j e c t i v e Decision prime  quality  impediment  of  decision to  reach  to  select adaptively  good,  though  not  Stone  (1987) a l s o  to  the  trade-off  weights  the  decision  evaluate  the  impact  cost,  i t will  maximisation predict  both  be  as  relatively  use  of  between maker  of  less c r i t i c a l .  complex  decision  places  and on  To  this  performance  end by  Cognitive  strategies.  examined  which l i n e to b e t t e r  in light  them.  of  the  empirical  understand the  well  trying  be  i n some cases be  a  well alter  the  the  relative  understand  and  accuracy  developed, decision  later  and  cognitive  which  aid,  c o n f l i c t i n g , can  r e s u l t s presented This  is a the  e f f o r t minimisation will  the  Changing  as  to  to  on  in  then  determine  should a s s i s t i n allowing  impact o f d e c i s i o n a i d s  In  effort.  cost  d e s i g n e d to reduce  hypotheses  o f r e a s o n i n g i s more d e f e n s i b l e . p r e d i c t and  In  as  individuals using  c h a p t e r 5. These hypotheses, which w i l l be  accuracy  to examine b o t h the  perspectives. of  effort  d e c i s i o n a i d s , which are  useful  types  that  a d e c i s i o n maker i s to minimise e x p e n d i t u r e s o f  i s viewed  a  necessarily  concluded  c o g n i t i v e c o s t s a s s o c i a t e d w i t h c e r t a i n d e c i s i o n s t r a t e g i e s may traditional  from  same time m i n i m i s i n g e x p e n d i t u r e s o f e f f o r t .  a v a i l a b l e evidence  of  makers  the  us  decision  process. The decision off.  next c h a p t e r w i l l aids  on  attempt  processing  to p r o v i d e some i n s i g h t i n t o the  which i s c o n t r o l l e d by  an  effort-accuracy  P r i o r to t h i s a n a l y s i s i t i s n e c e s s a r y to b r i e f l y address two 1) What i s the o v e r a l l v a l i d i t y o f the c o s t - b e n e f i t 2) Can  the p e r c e p t u a l  and  c o g n i t i v e views be  36  impact  approach?  reconciled?  of  trade-  issues.  3.4 Payne  (1982)  approach.  presented  framework.  The  task  adaptive  of  of  task  great  of  cost-benefit  more  as  which  the  basis  of  a  cost-benefit  i n c h a p t e r 2 i s a l s o most  argument. As  precisely  on  is  Further,  the  well,  nature  consistent  a great as  a  as  incorporated.  protocol  In  the  of  easily  review  such  with  this  complexity  data,  the  just  trade-offs cost-benefit  sense,  deal  comfort  rational  fact  such as  the  i n addition  of  decision  Payne  evidence  (1976),  to  to  cost-benefit face  of  provide  maker  (March,  if  the  ensures  that  the  argument has  validity.  those who  rational  alone  provides  f a c t t h a t they are a d a p t i n g s t r a t e g i e s  support  economically This  and  are aware o f the  empirical  man  similarity  conditions.  theory provides  interpreted  the  variables  to v a r i o u s  of  for  framework  simple f a c t t h a t s t r a t e g i e s change s y s t e m a t i c a l l y w i t h changes  evidence that subjects  belief  evidence  adaptivity  mechanisms.  deal  ample  l i t e r a t u r e presented  focusses  evidence  such  the  in light  which  provides  in  provides  Much o f  interpreted  V a l i d i t y o f the c o s t - b e n e f i t  Also  such  w i s h to p e r p e t u a t e 1977).  "cost  Actions  of  can  approach  a  the be  thinking"  cost-benefit  a  is will  have a l a r g e number o f advocates. One by  b a s i c d i f f i c u l t y w i t h the c o s t - b e n e f i t argument i s t h a t which i s f a c e d  any  namely  theory the  the c h o i c e problem,  claiming  problem o f  the  existence  infinite  of  a  regression  driving  (Einhorn  o f s t r a t e g y based upon c o s t - b e n e f i t how  provide  the  concern  that  is  this  most post  evaluation  critique  hoc  result  any  effort-accuracy  trade-off.  single  system,  decision  but  Payne  of  can  such be  (1982)  rather  there  37  and  controlling Hogarth,  Einhorn  and  explanations  interpreted  as  argues  that  there  are  multiple  a  (1981)  express  function is  If  decision  Hogarth and  a  routine,  1981).  i s i t s e l f considered  controlled?  complete  or  likely  interacting  the  of  an  not  a  systems.  Another it,  system may  being  Also, may  control  responsible  Johnson and  be  Payne  overcome  certain  for  high  the  level  execution  evaluation of  (1985) argue t h a t  i f we  conditions  the  speculate  i s a learned  that  the  the  the  or  strategy  problem o f  use  of  automatic b e h a v i o u r .  system which  Few  people  level. the  functions  would  In  this  these  respect  there  approach  strategies.  individuals  It  behave  in  parallels may  as  be  the  appears making  link  be  reasonably  decisions  implementation argument  of  while  particularly  Payne (1985j 1986) the  of  cognitive great  germane  to  the  s u p p o r t . I n essence i t can be p r e d i c t the way basis  directly  mechanism  as  e m p i r i c a l work o f Payne, B i g g s , Olshavsky and work o f Johns on and  and  has  study  of  to  such  the flaw  in  as  a  for  applied  invoking  model  the  making  by  the  the  actual  cost-benefit  and  is  not  decision  argued t h a t i f the c o s t - b e n e f i t concept h e l p s  i n d i v i d u a l s behave, then we  how  simulation  psychologists  decision  of  demonstrated  support  cognitive  describe  considering  Thorngate. In a d d i t i o n , which  theories  implementation  responsible  been  under  s p e c i f i e d manner.  to  robust  to  o t h e r s as w e l l as the  mechanisms  interest  i n the  regression  strategies  considered  a major c o n c e p t u a l  single to  infinite  Further,  i f i t were p e r f o r m i n g  argue  cost-benefit  decision  as  to  s e l e c t i o n program.  certain  which propose c o g n i t i v e mechanisms s h o u l d g e n e r a l l y be a  operate p a r a l l e l  to  have a u s e f u l e m p i r i c a l l y grounded  f o r d e v e l o p i n g d e c i s i o n support systems, a t l e a s t f o r a p a r t i c u l a r c l a s s  o f problems. Such a base i s n o t a b l y  absent from much e x i s t i n g work i n the  DSS  field.  3.5 Though the  the  invariance  Reconciling  cost-benefit in  the c o g n i t i v e and model  performance  perceptual  i s reasonably  across  38  a  wide  robust,  range  of  views i t does not  explain  tasks  in  cited  the  general  literature  on  e x a m i n a t i o n the two inspection  the  establishing processing, choice. problem  cognitive  of  structuring  the  model  in  in  judgments  activities  prospect  pair  theory.  does not  conclusion  that  to  a  there  solving  and  When  making  mechanism development  direct  such of  as  Tversky,  is a  a routine  need  surface  work  elementary  under  problems not  conducted  information  conditions  of  information  f o r by  binary  search  and  settings.  surround directly  in  The  filtering  comparable  prospect  and  to  the  theory.  In  to  choices  1979).  to  The  on  a  s i m i l a r to choice  between  o f an  computation  final  given that  comparisons  attribute  described  initial nor  multi-attribute  search  does  it  choose between p e r c e p t u a l  by  or  may  be  prospect  strategy i s lead  and  to  the  cognitive  Each would appear to operate a t d i f f e r e n t l e v e l s o f  possibly  consciousness.  of search  pairwise prospect  decision  on  studies  alternatives  cost-benefit  c o n t r o l over the c h o i c e  The  focus  accounted  generalize  imply t h a t the  approaches (Payne, 1982). problem  a  r e a s o n to expect t h a t the c o g n i t i v e mechanisms  of  executed u s i n g  attributable  On  multi-alternative  these  problems  and  automatically  not  to  generally  in  theory  a  tends  These problems are  (Kahneman  between  This  a l . , 1982).  reconciled.  multi-attribute,  r i s k y choice  problems  et  judgments most o f t e n  terms t h e r e would be no  described  be  approach  alternatives.  unidimensional,  choice  may  specific  structuring  majority  general  views  perceptual  making The  (Kahneman  approaches seem somewhat c o n t r a d i c t o r y ; however, on c l o s e r  two  the  heuristics  aids  Individuals  appear  s t r a t e g i e s and which items w i l l be  evaluation  of  theory  is  we  argue  may  preferences  invoked. that  In i t will  it  may  this be  be  sense,  to  compared. that for  39  hardwired.  a the  more f r u i t f u l  work i n areas such as s t r u c t u r i n g i n a m u l t i - a t t r i b u t e environment r a t h e r i n t r y i n g to change e v a l u a t i o n procedures which are  have  to  than  The  next  attribute, support providing  chapter  outlines  multi-alternative  those  processes  specific tools  both  some  strategies  settings  and  by complete  f o r decision  puts  forth  automation  t o augment human i n f o r m a t i o n  40  making i n m u l t i -  some  mechanisms  o f the procedure p r o c e s s i n g power.  to  and by  CHAPTER 4  - A BEHAVIOURAL D E C I S I O N THEORY APPROACH TO  4.0 The  preceding  complex, s t r a i n the  decision  different  chapters inducing  maker's  processing  strategies  Strategies  are  will  benefit  notion  augment  decision  makers.  choices  would  to  strategies proposed for  with  here  that  attaining this Typically  choice.  a  DSS  quantitative  input  is  on  to  basis  the  which  will  the  the  one  way  effort  probabilities support  of  processing.  importance,  influence  to is  good  of  cost-  This  cost-  which  truly  accuracy  required  (DSS)  to  accurate could  DSS  choices.  set of  accuracy.  improve  or  a  Under  different  final  a p p l i c a t i o n of and  a  been a t t r i b u t e d to  development  that  systems  of  final  implement  those  outcomes.  provide  a  It  is  mechanism  result.  DSS are  focusses generally  on  the  evaluation  applied  Problems are  to  f o r o b t a i n i n g an o p t i m a l  algorithm  are  multi-alternative  probabilistic.  c h a r a c t e r i s t i c s of p r e f e r e n t i a l choice  41  alternatives  i n that  prior  problems there  of  i s no  to a  known  s o l u t i o n or the v a r i a b l e s used In  optimal  preferential  of  semi-structured  semi-structured  i l l - s t r u c t u r e d p r o b l e m s w h e r e no  attribute,  decision  for  has  information  and  d e t e r m i n e d by  Presumably  decision  nature.  the  for  adopted  the  reduce  procedure or algorithm as  difficulty  t r a d e - o f f between e f f o r t  provide  higher  capacity  be  t o be  r u l e s which assess  be  this  informationload  benefit  may  Introduction  t a s k . Much o f  of  thought  DEVELOPMENT  have argued t h a t i n d i v i d u a l s f i n d d e c i s i o n making  limited  conditions  DSS  this  algorithm  choice  d i s s e r t a t i o n our exists,  problems.  problems i s t h a t the v a l u e  focus  namely m u l t i One  of  o f the  the input  data  i s generally reliable  and  d e t e r m i n i s t i c . In  this  sense we  w i t h problems o f d e c i s i o n making under c e r t a i n t y . However, due in  i n d i v i d u a l p r e f e r e n c e s , no  two  individuals w i l l  In  the  choice  literature  a  number  to d i f f e r e n c e s same  i n making a c h o i c e . There  t h a t can be taken to r e a c h a common s o l u t i o n  preferential  dealing  n e c e s s a r i l y make the  f i n a l c h o i c e nor w i l l they f o l l o w the same procedure are numerous paths  are  of  (Simon, 1982).  strategies  have  been  observed. Without possible  a  detailed  knowledge  of  an  individual's  f o r anyone to make a guaranteed  preferences,  i t is  not  s a t i s f a c t o r y c h o i c e f o r the d e c i s i o n  maker. A l s o t h e r e i s no g e n e r a l l y a c c e p t e d approach f o r i n f o r m a t i o n p r o c e s s i n g in  these  settings.  These  problems  differ  from  typical  s e t t i n g s s i n c e much o f the d a t a used i s q u a l i t a t i v e .  DSS  application  The normative model used  to s o l v e these problems i s p r e f e r e n c e or v a l u e t h e o r y which i s c l o s e l y to  expected  significant One  to d e v e l o p i n g  extract  proceed  estimates  to suggest  literature  examples  of  has  however,  as  in  such  DSS  of  f o r such  we  this  area  preferences  systems are  dating  have  indicated  i n r e a l world  from  back  d e s c r i b e d by  the  and  (1961).  and  More  McFadden  recent  (1980)  and  (1987).  r e a c t i o n o f u s e r s to f o r m a l systems f o r s u p p o r t i n g p r e f e r e n t i a l  been mixed and  then  There i s a g r e a t d e a l  Edwards  Humphreys  settings.  implement models  d e c i s i o n maker  estimates. to  previously,  problem  a p p l i c a t i o n s i s to  c h o i c e s based upon those  Humphreys and Wisudha The  theory;  d e v i a t i o n s from t h i s model occur  approach  which  of  utility  related  t h e i r use  has  been f a r from widespread  (Dickmeyer,  choice 1984).  John e t a l . (1983) found t h a t such systems were judged by u s e r s as i n f e r i o r to an not  analyst performing  the  d i f f e r e n t between the  same t a s k , system and  42  though the analyst.  quality  In t h i s  of  the  outcomes  was  r e s p e c t systems which  utilize  s o p h i s t i c a t e d techniques f o r e l i c i t i n g and computing p r e f e r e n c e s  upon u t i l i t y also  t h e o r y may  points  decision that  to  the  analysis  the  accepted  because appeal  supported  by  reluctance  of  ( C h r i s t e n and  solutions  intuitive  have low user acceptance.  they and  the  are are  fact  by  will  from  difficult  that  to  accept  analytic  an  experimental  Fischoff  the  techniques  choice  These  models  suggested not  be  process  with  no  seem t o  be  found  little  use  Another p o s s i b l e e x p l a n a t i o n f o r the  be  of  may  arguments  have  evidence  outcomes  (1981) has  incomprehensible  to j u s t i f y .  formal  a p p l i c a t i o n o f such models may  argued  decision  output  p r a c t i c a l management s i t u a t i o n s .  explicit  individuals  Somet, 1980).  suggested  Other  based  in  limited  t h a t the c o g n i t i v e c o s t a s s o c i a t e d w i t h  the  i d e n t i f i c a t i o n o f p r e f e r e n c e f u n c t i o n s i s q u i t e h i g h . Keen (1979) has  t h a t i f the p e r c e i v e d b e n e f i t o f the t o o l does not exceed  not  be  used.  This  may  be  one  of  the  difficulties  i t s cost, i t  inherent  i n complex  systems t h a t perform p r e f e r e n c e d e t e r m i n a t i o n . Given  that  w i d e l y used  tenets  be of  individual  and  utility  modelling  or p a r t i c u l a r l y s u c c e s s f u l ,  are developed would  preference  of  t o support interest.  DSS,  that  maker.  be  systems the  The  would  system  conform  should  introduction  of  have  not  that other techniques  augment the a c t u a l b e h a v i o u r  These  namely  decision  and  i t may  approaches  to  been which  o f d e c i s i o n makers one  support,  of not  limited  the  central  replace  support  the  models  in  these areas can be compared to the use of DSS  i n problem domains where o p t i m a l  management  s u c c e s s f u l implementation  s c i e n c e models  and Scott-Morton, In  different  as  to  achieve  (Keen  1978).  applying decision  techniques  failed  preference  alternatives  strategies  very  evaluation  (see, Biggs,  or  1978  43  few  people  expected and  explicitly value  Troutman and  when  employ  such  evaluating  Shanteau,  1976  for  evidence with respect  to p r e f e r e n t i a l c h o i c e and Kahneman e t a l . (1982) f o r  overview o f r e s u l t s o f p r o b a b i l i s t i c e v a l u a t i o n s ) . section  c h o i c e s between p a i r s  mechanism  such  mechanisms are regard  use,  The  described  evaluation  by  at  Thus  is  thought  i t might  this  to  seem  a i d s which would reduce the  prospect  and  s e l e c t i o n of strategy  however,  strategy.  that  risky alternatives  presumably h a r d w i r e d  supporting  endeavour. and  as  of  As  be  to  may  likely  Furthermore,  t a s k demands.  be  reasonable  by  that  a  a  the  previous  governed by  less  conscious  a  such  In  than  for structuring information controlled  cognitive  are  theory.  invariant level  argued i n the  an  this  fruitful  acquisition cost-benefit  development  of  l o a d o f implementing v a r i o u s  decision  strategies  c o u l d encourage more thorough a n a l y s i s .  4.1 This  section  descriptive choice  is  models  problems.  directly  to  decision  aid  setting. w i l l be Four  the  (Payne,  5  intended  types  designed  of to  1976;  support  For  (AD), 1978;  in  that  which  decision which  formal  are  definitions  considerations  employed  for  need  terms,  these  to  be  models  will  incorporated a  basic  in preferential  making  in  can  drawn from  be  some  point into  preferential this  a  choice  analysis  4.2.  strategies  Biggs,  search  mechanisms  commonly  p r e f e r e n t i a l choice  difference  describe,  formal  presented i n section decision  to  information  Providing  Some d e s i g n  alternative, additive  of  Models o f p r e f e r e n t i a l c h o i c e  employed  problems a r e : 5  conjunctive  (CONJ) and  Biggs  a l . , 1985;  et  s i m p l i c i t y these w i l l now  problems." 44  be  in  multi-attribute,  additive  compensatory  e l i m i n a t i o n by Jarvenpaa,  r e f e r r e d to as  multi-  aspects  1988).  (AC), (EBA)  Each  " p r e f e r e n t i a l choice  of  these  e n t a i l the  carry  out  Utilising  an  use  of various  examination  useful  to  information generally  additive strategies  processes  each can (EIPs)  to  be  typically  from  the  is  in  some sense  outcomes.  However,  making  that  is  capabilities  and  between  to  them.  l o a d on  the than  EBA).  for  EIPs  memory at  are  example  a  (Johnson the  i d e n t i f y the  processes single  points,  odds w i t h  milli-second  the  more  which  shift  Payne,  in  we  The  These idea  they  failing do  not  limitations  of  many  directly  of  the  in solving  s u p p o r t based upon d e s c r i p t i v e models, we  normative take  may  behind  problems.  t r a d i t i o n a l normative  decision  are  s u p p o r t mechanisms based upon  approach  d e c i s i o n makers s h o u l d behave i n o r d e r to o b t a i n the  would  attention,  1985).  level.  d e c i s i o n makers a c t u a l l y behave at  elementary  or r e t r i e v i n g an element o f  and  s p e c i f i c EIPs i s to t r y to d e r i v e  which s p e c i f i e s how  choice  (AC,AD) b e i n g more d i f f i c u l t  implemented and  comparing two  term  u n d e r s t a n d i n g o f how  This  a  i n order  l e v e l s of cognitive  (CONJ,  involved.  p r o c e s s e s which o c c u r  evaluating an  long  be  atomic,  reading a s i n g l e data point, information  make  operations  mechanisms f o r s u p p o r t i n g these types o f s t r a t e g i e s i t  review how  consider  and  varying  e l i m i n a t i o n or f i l t e r i n g s t r a t e g i e s In o r d e r to d e r i v e  is  alternatives  these s t r a t e g i e s w i l l p l a c e  d e c i s i o n maker, w i t h the the  of  elementary c o g n i t i v e  approaches  into  account  maker.  By  desired  to the  trying  decision specific to  overcome t h i s problem and  build derive  systems which a l l e v i a t e the problems t y p i c a l l y f a c i n g a d e c i s i o n maker such as l i m i t e d short The  term memory and  models can be  r e l a t i v e l y slow s e r i a l p r o c e s s i n g  e v a l u a t e d and  c l a s s i f i e d a l o n g two  1)  compensatory v e r s u s noncompensatory  2)  independent v e r s u s dependent  basic  speed. dimensions  and  Models which e v a l u a t e a s i n g l e a l t e r n a t i v e a t a time a l l o w i n g  45  low  ratings  on  one  attribute  be  compensatory.  elements  do  to be  Those which  not  attributes;  permit  these  prototypical constant  of  is  major a  analysis  and  by is  alternative  of  by  high  i s manifested searched  filtering  comparisons  of  each  searches the  or  are  alternative  and  before  moving  on  to  across  The  occur  s a i d to be  in  has  been  established.  information  the  next  is  other  use fact  of  a  that  a  resulting  in  decision-making case,  implies  until  an  Independence the  basic  examined  Only  one  independent are  comparing  on  alternatives  i.e.,  one.  specific  alternative.  this  across  alternatives,  the  between  made  which are not  The  to  alternative.  Independence,  all  said  values  elimination  interdimensional,  are for  each  not  alternative  which  by  for  discriminating  are  searches which searches are  for  independence.  Strategies  attribute  unit  for  a  is of  given  alternative  is  demonstrated  by  values.  Such  intradimensional.  a d d i t i v e - c o m p e n s a t o r y model  Analysis  using  would examine one the  evaluation  the  evaluation  sum  offset  levels  noncompensatory.  is  imply  e v a l u a t e d a t a time.  product  be  strategy  criteria  form  measure  suggested  4.1.1  to  information  other a t t r i b u t e s  minimum t h r e s h o l d  considered  search  strategies  second  approaches  analysis  values  on  amounts o f i n f o r m a t i o n b e i n g s e a r c h e d f o r each  The  overall  high ratings  require  are  compensatory  Noncompensatory  that  low  models  amount  variable  o f f s e t by  of  would  the be  the  Additive-Compensatory  alternative  score  for  would be each  a t t r i b u t e v a l u e and chosen.  implies  a t a time a l o n g a l l r e l e v a n t  each a t t r i b u t e a  Model  Thus,  assigned  alternative weight; the  46  would  the  a  subject  attributes.  During  a weight. be  that  After  derived  alternative  with  Additive-Compensatory  completing summing the  the  highest  strategy  is  compensatory; values  a l l attributes  on some a t t r i b u t e s  are evaluated,  weighted and summed a l l o w i n g  t o be o f f s e t by h i g h v a l u e s  on o t h e r  low  attributes.  S i n c e each a l t e r n a t i v e i s examined i n i t s e n t i r e t y and no comparisons a r e made until  an  overall  independence.  evaluation  has  I n terms o f c o g n i t i v e  been  arrived  at,  the  model  exhibits  l o a d t h i s i s a r e l a t i v e l y demanding model  f o r the d e c i s i o n maker t o employ w i t h o u t the use o f any form o f d e c i s i o n It  involves  the use o f a l l a v a i l a b l e  is associated  i n f o r m a t i o n and c o n s i d e r a b l e  aids.  processing  w i t h each item o f i n f o r m a t i o n examined.  4.1.1.1 Formal d e s c r i p t i o n o f s t r a t e g y More f o r m a l l y as  the o p e r a t o r s used t o invoke t h i s model c o u l d  be  described  follows: Move t o a l t e r n a t i v e j a t random from the c a n d i d a t e L00P1 I f number o f n o n e v a l u a t e d a t t r i b u t e s  set.  f o r chosen a l t e r n a t i v e =  number o f n o n e v a l u a t e d a l t e r n a t i v e s >  and <ji  move t o next a l t e r n a t i v e j+1  LOOP 2 I f the number o f n o n e v a l u a t e d a t t r i b u t e s f o r the chosen a l t e r n a t i v e > 9 move t o the next a t t r i b u t e i r e a d the v a l u e o f the a t t r i b u t e a  i J  r e t r i e v e the weight (w ) a s s o c i a t e d A  (attribute i f o r alternative j ) w i t h the g i v e n a t t r i b u t e i  combine the a t t r i b u t e v a l u e and weight g i v i n g v ^ a ^ r e t r i e v e current add  total  (Sum(w a )) i  ij  the weighted v a l u e t o t o t a l f o r c u r r e n t  s t o r e new t o t a l  47  alternative  Go t o LOOP 2 Go t o LOOP 1  alternatives = 0  I f number o f nonevaluated  r e t r i e v e s c o r e s f o r each a l t e r n a t i v e ( j ) compare a l l s c o r e s s e l e c t the a l t e r n a t i v e w i t h h i g h e s t s c o r e  This the  i s a minimal c o m p l e x i t y  existence  retrieved. that  o f a preference  The weights  f o r each  r e l a t i v e l y demanding.  v e r s i o n o f the AC model structure  (Keeney  are stored i n long  attribute  a  retrieval  Storage  6  must  i t assumes  and R a i f f a ) which  term memory take  i n that  place;  (LTM) which such  c a n be implies  a c t i v i t i e s are  o f a u n i t o f i n f o r m a t i o n i n t o LTM takes on the  o r d e r o f t e n seconds; r e t r i e v a l i s a m i l l i s e c o n d o p e r a t i o n b u t i s s u b j e c t t o a relatively exist  high  prior  retrieved  to  error  rate.  analysis,  many times  For t h i s  would  have  model t o be  attribute created  weights, and  unless  stored  once  d u r i n g the e v a l u a t i o n o f a s e t o f a l t e r n a t i v e s .  they and  Also a  r u n n i n g t o t a l o f the s c o r e f o r a g i v e n a l t e r n a t i v e would have t o be m a i n t a i n e d and updated as would a v e c t o r o f f i n a l s c o r e s f o r each a l t e r n a t i v e . The v a l u e s o f the a t t r i b u t e s cost  a r e r e a d from an e x t e r n a l source  which i s a r e a s o n a b l y low  f u n c t i o n , t a k i n g on the o r d e r o f 0.3 seconds.  Since a l l information i s  I n t h i s case the s c o r e i s a p r e f e r e n c e v a l u e and the h i g h p o s i t i v e v a l u e r e l a t e s t o the t o t a l p r e f e r e n c e s c o r e . I t i s e n t i r e l y p o s s i b l e t h a t t h i s high preference s c o r e was a r r i v e d a t by the a g g r e g a t i o n o f v a l u e s on a t t r i b u t e s where lower v a l u e s would r e s u l t i n a h i g h e r l e v e l o f p r e f e r e n c e . I n l a t e r model the > s i g n i s used t o i n d i c a t e t h i s same p r e f e r e n c e r e l a t i o n s h i p . Thus e x p r e s s i o n such as a ^ > c^ s h o u l d be r e a d as a i s p r e f e r e d t o c r a t h e r than a i s g r e a t e r than c s i n c e i n some cases s m a l l e r v a l u e s w i l l be p r e f e r a b l e to l a r g e r ones. 6  48  examined using important.  this  strategy, the  issue of  ordering a l t e r n a t i v e s  i s not  Combining the a t t r i b u t e value and weight i s g e n e r a l l y assumed to  be a m u l t i p l i c a t i v e operation (however, any method of v a r y i n g complexity could be  used).  The  current t o t a l value f o r the a l t e r n a t i v e must be  (again, presumably from LTM)  and the new  take on the order of 0.3 - 0.5 seconds).  retrieved  value added to i t (such  processes  The new t o t a l i s then stored i n LTM.  The c o n t i n u a l summation of the updated t o t a l minimizes the number of accesses to memory and makes t h i s a reasonably e f f i c i e n t form of the AC model. set  of  instructions  alternative  until  will  be  executed  i t i s completely  for  each  evaluated,  attribute  at which time  of  This  the  given  attention i s  s h i f t e d ( v i a a MOVE command) to the next a l t e r n a t i v e . When a l l a l t e r n a t i v e s have been evaluated, i t i s necessary to r e t r i e v e the t o t a l f o r each a l t e r n a t i v e . of  this  The scores must be compared.  mechanism would be  dependent upon the number of a l t e r n a t i v e s .  minimal cost comparison could be conducted short  term  memory (STM)  The p r e c i s e nature A  i f a l l scores could be read into  f o r processing. The  simplest approach may  r e t r i e v e two scores, compare them, r e t a i n the highest and r e t r i e v e  be  to  another.  This process would continue u n t i l a s i n g l e a l t e r n a t i v e remains. A l t e r n a t i v e l y , i t i s conceivable that on completing the e v a l u a t i o n of a l t e r n a t i v e n+1,  i t is  compared d i r e c t l y to the t o t a l f o r the previous a l t e r n a t i v e n and the b e t t e r score i s r e t a i n e d . This procedure pure AC model, but would be reduce  some  of  the  would not be t o t a l l y c o n s i s t e n t w i t h the  a p r a c t i c a l means by which a d e c i s i o n maker could  storage  and  processing  demands  a s s o c i a t e d with  the  implementation of t h i s strategy. In a l l of t h i s a n a l y s i s we are assuming that the d e c i s i o n maker does not have recourse to any problem s o l v i n g process  either  e x t e r n a l support  f o r the  i n terms of e x t e r n a l storage mechanisms or 49  t o o l s to support  the p r o c e s s i n g and  4.1.1.2 S u p p o r t i n g Given  an  the  i n t e g r a t i o n of  information.  strategy  algorithm  for  processing  in  such  a  manner  i t i s possible  s p e c i f y s e v e r a l components o f the d e c i s i o n procedure which c o u l d be If  the  vector  of  weights  mechanism to p l a c e be  a relatively  exactly  what  formalise  and  certainty. processing  value  is  function a  also  strategy  c o u l d e a s i l y be variety  scores to  develop the  o f weights  scores and  i s obtained  select  implement  and  and  more  However, a l l o f  these  and  i t would  alleviate  support  limited  support  and  be  a  the  under  for  storage  do,  this  demands  reading of a t t r i b u t e s into o f weights and  provided  addition of a  to  trade-offs  i n a computer-based d e c i s i o n a i d .  values  with  This i s  attempts  f o r making  However, s t o r a g e  the h i g h e s t v a l u e . would  1976)  computational  A t t e n t i o n to and  d e c i s i o n maker.  accommodated  Raiffa,  structure  to  alleviating  and  o f b a s i c mathematical f u n c t i o n s c o u l d be  combination total  (Keeney  possible  by  the  a v a i l a b l e , along  a common s c a l e o f measurement, i t would  preference  p l a c e d on the d e c i s i o n maker. must remain w i t h  supported.  t a s k to automate the e n t i r e d e c i s i o n p r o c e s s .  implement It  a l t e r n a t i v e was  a l l a t t r i b u t e s on  simple  a  f o r each  to  simple  totals  Similarly,  to  Once a v e c t o r sort  the  computational  mechanisms presuppose an  ability  of  outcome  These f u n c t i o n s are a l l r e l a t i v e l y d e c i s i o n maker's  a  f o r s p e c i f y i n g the  totals.  matter  STM  easy  burden.  to weight  the  v a r i o u s a t t r i b u t e s , something which i s seldom done i n an e x p l i c i t manner. Even  in  the  case  where  weights  are  unavailable,  some  basic  support  mechanisms c o u l d be  developed; namely, simple  rank o r d e r i n g o f a t t r i b u t e s and  of  facilitate  choice.  a l t e r n a t i v e s to  display,  an  i n d i v i d u a l may  e v a l u a t i o n and  Also  i n terms o f  perform b e t t e r when each a l t e r n a t i v e i s  50  data  presented  i n d i v i d u a l l y to avoid general, without  the p o s s i b i l i t y o f d i s t r a c t i o n from i r r e l e v a n t d a t a .  f o r c i n g a s p e c i f i c a t i o n of a value  function  there  In  is little  t h a t c a n be done t o support t h i s mode o f a n a l y s i s .  4.1.2 The a d d i t i v e - d i f f e r e n c e The  Additive-Difference  individual  makes  alternatives. results  These  model  model  comparisons  comparisons  as a d d i t i o n  (Tversky,  between  1969)  specific  are evaluated  f o r each  i s used  comparisons  attribute  and the  differences  retained  while  compared a g a i n s t  will  alternatives  the o t h e r  i s complete,  i s rejected.  The  dominant  retained  This  is a  p e r m i t low  The model i s n o t  comparisons a r e made between a l t e r n a t i v e s . f o r two  an  f o r two  t o produce an o v e r a l l e v a l u a t i o n . o f the weighted  when  attributes  t o be o f f s e t by h i g h e r v a l u e s on o t h e r a t t r i b u t e s .  independent s i n c e the  (AD)  a r e weighted and summed  compensatory values  pairwise  model  When one s e t o f alternative i s  alternative  i s then  the next a l t e r n a t i v e .  4.1.2.1 Formal d e s c r i p t i o n o f s t r a t e g y A  detailed  description  o f the EIPs  involved  i n using  follows: Move  t o an a l t e r n a t i v e j ( a t random) from c a n d i d a t e s e t  Drop j from c a n d i d a t e s e t S t o r e j as c u r r e n t - p r e f e r r e d - a l t e r n a t i v e j * LOOP 1 I f c a n d i d a t e s e t > 0 Move t o a l t e r n a t i v e k from c a n d i d a t e s e t Drop k from c a n d i d a t e s e t Move t o a t t r i b u t e i  51  the AD model i s as  LOOP 2 I f a t t r i b u t e s e t > 0 read a t t r i b u t e value a  t J  read a t t r i b u t e value a  i k  subtract  a  from a  Li  i k  giving difference  (d ) ±  r e t r i e v e weight (w ) f o r a t t r i b u t e ( a ^ t  combine d add  wd ±  i  and w  x  giving w d i  i  to t o t a l difference  A  score 0  w^  s t o r e new t o t a l d i f f e r e n c e fi w ^ move t o next a t t r i b u t e i + 1 Go t o LOOP 2.  Retrieve t o t a l  difference  If total difference eliminate store  > 0  alternative j  alternative k i n current-preferred  else eliminate  alternative k  Go t o L00P1. S e l e c t j * as d e s i r e d  Like control to  t h e AC model routines  alternative  this  model  assumes  t o monitor the s t a t u s  that  o f the a n a l y s i s  a c t u a l l y make c h o i c e s between a l t e r n a t i v e s .  outcome o f a n a l y s i s alternative alternatives  a r e minimal s i n c e  i s being  carried.  c a n be d i s r e g a r d e d .  includes  Contrast  a set of  and e x e c u t i o n  Storage requirements  a t any time o n l y t h e c u r r e n t  Information  52  processing  about this  previously  with  the AC  routines f o r the preferred  eliminated model  which  requires until  the s t o r a g e  o f each  outcome  since  a l l i n f o r m a t i o n has been e v a l u a t e d .  no comparison  i s f o r m a l l y made  Thus from a c o n t r o l p o i n t o f view  t h i s model i s l e s s complex than the AC model.  4.1.2.2 S u p p o r t i n g the s t r a t e g y Computational  support  would  between the v a l u e s f o r each a , ti  difference  function  represented necessary.  in a  simply  a  i k  a  function  t o take  the d i f f e r e n c e  p a i r . I f a l l a t t r i b u t e s a r e numeric,  involves  qualitative  Other f a c i l i t i e s  involve  subtraction.  form,  a  more  If  complex  would be e s s e n t i a l l y  the  variables  mechanism  terms  the same as those  alternatives  o f i n f o r m a t i o n d i s p l a y i t may be a p p r o p r i a t e shown  at  any  time.  A  a l t e r n a t i v e would l i k e l y be b e n e f i c i a l . task  structuring  function.  could  Another  be  provided  facility  to  of totals.  the e x p l i c i t  use  of  general  representation of a p r e f e r e n t i a l  the  ability  t o move  single  useful  rows  o r columns  i n comparing a l t e r n a t i v e s .  e s p e c i a l l y important a  even  screen. without  a  value  s t r u c t u r i n g mechanism which may be u s e f u l i n implementing  the  on  dominated  i n terms o f  t h i s s t r a t e g y would be the a b i l i t y t o move a l t e r n a t i v e s t o g e t h e r .  involved  o n l y two  a  A g a i n the minimal support without  be  i n the AC  t o have  eliminate  are  might  model i n terms o f s t o r a g e and r e t r i e v a l o f weights and a c c u m u l a t i o n In  this  would  c h o i c e problem likely  ease  Considering  i n matrix  some  form,  o f the s t r a i n  I n a computer-based system t h i s would be  i f a l t e r n a t i v e s i , j c o u l d n o t be s i m u l t a n e o u s l y d i s p l a y e d Thus, recourse  basic  task  structuring  t o implementing  value  support  aids  may  f u n c t i o n s which  prove a c t to  f o r m a l i s e the p r e f e r e n c e s t r u c t u r e . The they  previous  two models, AC and AD, can be c h a r a c t e r i z e d by the f a c t t h a t  l e a d t o a complete a n a l y s i s o f any i n f o r m a t i o n s e t .  53  AD i s s l i g h t l y  less  demanding Support  to  for  both  preferences Although  implement  in  structuring encourage  those  order  some  since  these  models  to  basic  the  make  are  storage  involves value  elements  can  trade-offs be  to  more  elimination  strategies  We by  will  now  aspects.  structures  and  that  are  Neither  of  these models assumes the  complete s e a r c h o f the problem space.  this  nor  As  of  values. task  possible  to  analysis  by  adopted  models, c o n j u n c t i v e  integration  assessing  support be  typically  large.  to  pare  analysis.  two  of  as  for  may  i t may  conducting  to  purposes  which  time  turn  for  elimination  not  aggregation  However,  down l a r g e problem spaces p r i o r to d e t a i l e d  are  techniques  introduced  minor.  spend  requirements  developing  relatively  individuals  facilitating  its  do  use they  of  and  elimination,  detailed  generally  a r e s u l t they are  preference  result  in  thought to be  a  less  c o g n i t i v e l y demanding s t r a t e g i e s .  4.1.3  The The  c o n j u n c t i v e model conjunctive  some minimum fails  independently evaluation found  a  be  of  level  the  dropped  until  specifies  it  from is  which  satisfies are  This  a l l attributes.  l e v e l s set  either  Evaluation  are  evaluated  Any  alternative  of  from  This  makes the  to f i n a l  excluded  from  is  which will  evaluated or  the  alternative order  selection.  is  i n which  This  model  below t h r e s h o l d s  further  models.  against  attributes  consideration  stops when an  a l l filtering  54  the  alternative  a l t e r n a t i v e s which are  immediately  c h a r a c t e r i s t i c of  Each  generally  a l l minimum c r i t e r i a .  are  f o r any  eliminated  s e a r c h e d extremely important  attribute  is a basic  alternatives  consideration.  o b v i o u s l y noncompensatory s i n c e particular  that  along  threshold  i s completed.  alternatives is  acceptance  to meet one  automatically  model  on  consideration.  Each e v a l u a t i o n  is  independent; assumes  no comparisons  that  the t h r e s h o l d  a r e made between a l t e r n a t i v e s . values  are e x t e r n a l l y  This  generated  independence and  dependent upon v a l u e s o f a c t u a l a l t e r n a t i v e s c o n t a i n e d i n the c h o i c e  4.1.3.1 D e s c r i p t i o n More  are not set.  o f the s t r a t e g y  formally,  analysis  under  a  conjunctive  model  would  proceed  follows: LOOP 1 I f number o f n o n e v a l u a t e d a l t e r n a t i v e s = 0 reset threshold  values (c)  s t o r