ATTRACTION EFFECT: THE INFLUENCE OF DECOY AND ‎ PHANTOM CANDIDATES ON JOB-FINALIST

(1)

ATTRACTION EFFECT: THE INFLUENCE OF DECOY AND ‎ PHANTOM CANDIDATES ON JOB-FINALIST

CHOICE

BY

AQBAL HAFIZ IZUDDIN BIN ABD HAMID

A dissertation submitted in fulfilment of the requirement for the degree of Master of Human Sciences in Psychology

(Industrial and Organisational)

Kulliyyah of Islamic Revealed Knowledge and Human Sciences

International Islamic University Malaysia

MAY 2018

(2)

ii

ABSTRACT

There is strong evidence which suggests that the task of choosing a single job candidate from a small set of comparable finalists can be influenced by the contextual factor of attraction effect. The present study examined the influence of decoy and phantom candidates on job-finalist choice. A total of 150 participants were selected and they role played as hiring managers in a simulated employee selection scenario.

Results from McNemar’s chi-square revealed the following: (i) there is significant difference in the number of participants who chose the target candidate in control and decoy conditions at p < .001, (ii) there is no significant difference in the number of participants who chose the target candidate in control and phantom conditions at p >

.05 and (iii) there is significant difference in the number of participants who chose the target candidate in decoy and phantom conditions at p < .05. In addition, results from further analysis on participants with different backgrounds showed the following: (i) the effect of a decoy candidate is significant on personnel selection students at p < .05, but not on personnel selection workers and laypeople at p > .05, (ii) the effect of a phantom candidate is not significant on all three subgroups at p > .05 and (iii) the difference between the effects of decoy and phantom candidates is significant for personnel selection workers at p < .05, but not for personnel selection students and laypeople at p > .05. The implications of these findings were discussed, and recommendations for future research were provided.

(3)

iii

صخلم ثحبلا

ABSTRACT IN ARABIC

كانى ةلدأ ةيوق يرشت لىإ نأ ةمهم رايتخا حشرم

ةفيظو ةدحاو نم

ةعوممج ةيرغص

نم

ينلىأتلدا ينيئاهنلا

ثيح نكيم نأ رثأتت لماوعلبا ةيقايسلا

يرثأتل بذلجا . دقل تلوانت ةساردلا

ةيلالحا يرثتأ تاحشرلدا ةفدهتسلدا

ةلوهلمجاو ىلع

رايتخا لمعلا

يئاهنلا . تمو رايتخا وعوممج

نمضتت اًكراشم 051

أو اود مىرود ب و مهفص نيريدم

ينيعتلل في

ويرانيس ةاكامح

رايتخا

فظولدا . دقل تفشك جئاتنلا

تيلا تم لوصلحا اهيلع

نم ةحاس "

رامينكام "

ام يلي ( : ) 0

كانى فلاتخا يربك

في ددع ينكراشلدا نيذلا

اوراتخا حشرلدا

هتسلدا فد في ةرطيسلا طورشو

معطلا في

p < .001

، ( ) 2 لا دجوي قرف يربك في ددع ينكراشلدا نيذلا

اوراتخا حشرلدا

فدهتسلدا في

ةرطيسلا فورظلاو

ةيهمولا دنع

p > .05

و ( ) 3 كانى قرف يربك في ددع

ينكراشلدا نيذلا

اوراتخا حشرلدا

فدهتسلدا في

فورظ رشلا

حبشلاو في

.p < .05

ةفاضلإبا لىإ

،كلذ ترهظأ جئاتنلا

دعب ديزم نم ليلحتلا ينكراشملل

يوذ تايفللخا ةفلتخلدا

ام يلي ( : ) 0

يرثتأ حشرم ةبلاطلا ةيهمأ تاذ

ىلع بلاط رايتخا

ينفظولدا في

p < .05

، نكلو سيل ىلع

لامع رايتخا ينفظولدا سانلاو

ينيداعلا دنع

p > .05

، يرثتأ

(2)

حشرلدا ةيهمولا

سيل ايربك في

عيجم تاعوملمجا ةيعرفلا

ةثلاثلا دنع

p > .05

و ( ) 3 قرفلا ينب رثاآ معطلا ينحشرلداو ةيهمولا

وى رمأ مهم ةبسنلبا ينلماعلل

في رايتخا ينفظولدا

في

p < .05

، نكلو سيل بلاطل رايتخا

دارفلأا سانلاو ينيداعلا

في

.p > .05

تتم ةشقانم رثالآا

ةبتترلدا ىلع هذى جئاتنلا

، تمدقو

تايصوت ثوحبلل

في لبقتسلدا

.

(4)

iv

APPROVAL PAGE

I certify that I have supervised and read this study and that in my opinion; it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Master of Human Sciences in Psychology (Industrial and Organisational).

………..

Alizi Alias Supervisor

………..

Harris Shah Abd Hamid Co-Supervisor

I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Master of Human Sciences in Psychology (Industrial and Organisational).

………..

Jusmawati Fauzaman Examiner

This dissertation was submitted to the Department of Psychology and is accepted as a fulfilment of the requirement for the degree of Master of Human Sciences in Psychology (Industrial and Organisational).

………..

Shukran Abd Rahman

Head, Department of Psychology This dissertation was submitted to the Kulliyyah of Islamic Revealed Knowledge and Human Sciences and is accepted as a fulfilment of the requirement for the degree of Master of Human Sciences in Psychology (Industrial and Organisational).

………..

Mohammad Abdul Quayum Dean, Kulliyyah of Islamic Revealed Knowledge and Human Sciences

(5)

v

DECLARATION

I hereby declare that this dissertation is the result of my own investigation, except where otherwise stated. I also declare that it has not been previously or concurrently submitted as a whole for any other degrees at IIUM or other institutions.

Aqbal Hafiz Izuddin Bin Abd Hamid

Signature………....………. Date …….……….

(6)

vi

COPYRIGHT

INTERNATIONAL ISLAMIC UNIVERSITY MALAYSIA

DECLARATION OF COPYRIGHT AND AFFIRMATION OF FAIR USE OF UNPUBLISHED RESEARCH

ATTRACTION EFFECT: THE INFLUENCE OF DECOY AND

‎

PHANTOM CANDIDATES ON JOB-FINALIST CHOICE

I declare that the copyright holder of this dissertation are jointly owned by the student and IIUM.

No part of this unpublished research may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise without prior written permission of the copyright holder except as provided below

1. Any material contained in or derived from this unpublished research may be used by others in their writing with due acknowledgement.

2. IIUM or its library will have the right to make and transmit copies (print or electronic) for institutional and academic purposes.

3. The IIUM library will have the right to make, store in a retrieved system and supply copies of this unpublished research if requested by other universities and research libraries.

By signing this form, I acknowledged that I have read and understand the IIUM Intellectual Property Right and Commercialization policy.

Affirmed by Aqbal Hafiz Izuddin Bin Abd Hamid

……..……….. ………..

Signature Date

(7)

vii

DEDICATION

This dissertation is dedicated to my beloved late father, Almarhum Hj. Abd. Hamid bin Hj. Mansor,

and my beloved late mother, Almarhumah Hjh. Minah bte Hj. Piah.

(8)

viii

ACKNOWLEDGEMENTS

In the name of Allah, the Most Beneficent, the Most Merciful. All praise is due to Allah, the Lord of the Worlds. Peace and blessings be upon His beloved, Prophet Muhammad.

First of all, I would like to show my heartfelt appreciation to my deeply- missed late father, Almarhum Hj. Abd. Hamid bin Hj. Mansor, for raising me with advice, discipline, guidance and teachings. My heartfelt appreciation also goes to my beloved late mother, Almarhumah Hjh. Minah bte. Hj. Piah, for nurturing me with love, care, patience and humour. Both of them have made me who I am today as a person. Not to forget, my sister, Mrs. Natasha Hani Abd Hamid for providing me accommodation and utilities throughout the administration of the questionnaire, and all other siblings who supported me throughout the process.

Secondly, I would like to thank my supervisor, Dr. Alizi Alias for his guidance, tolerance, and patience throughout the whole process of writing this dissertation. His supports transcend his working hours. In addition, he always gave me direction and put full trust in me in completing this research. I also would like to thank my co-supervisor, Dr. Harris Shah Abd Hamid, for his suggestion to expand my samples to include three subgroups so that my research would become more publishable, Dr. Jusmawati Fauzaman for her lecture and mentorship on how to plan and conduct this research, Br. Ferdaus Harun for conducting very helpful sharing sessions on his knowledge and experience completing his Master’s research, Br.

Haikal Pua’ad for helping me to identify the correct statistical analyses and guiding me to write this dissertation, and Mrs. Nurul Sharmila Idayu Nizan for proofreading this dissertation.

Thirdly, I feel obliged to show courtesy to the personnel selection workers, personnel selection students, and laypeople participating in this research despite their hectic schedules. They gave their full cooperation for this research. In particular, I have gained a lot of insights from them on how to improve this research and they even helped me to distribute the questionnaire.

Lastly, I am indebted to all my lecturers, family, relatives, friends, and acquaintances that have helped me directly and indirectly throughout completing this research. Their presences alone have given me strength in completing it, not to mention their assistance.

Forgive me if I forget to include rightful names in this writing and for any shortcoming. May Allah recompense for your good deeds.

(9)

ix

LIST OF TABLES

Table ‎3.1 Sequences in the Presentation of Conditions 26 Table ‎3.2 Control Condition Candidate Assessment Scores 28 Table ‎3.3 Decoy Condition Candidate Assessment Scores 29 Table ‎3.4 Phantom Condition Candidate Assessment Scores 29 Table ‎4.1 Frequency and Percentage of Participants based on Subgroups 37 Table ‎4.2 Participants’ Choice Percentages Distribution across Conditions 38 Table ‎4.3 Choice Percentages by Control-Decoy Condition 39 Table ‎4.4 Choice Percentages by Control-Phantom Condition 40 Table ‎4.5 Choice Percentages by Decoy-Phantom Condition 41 Table ‎4.6 Choice Percentages by Subgroup on Control-Decoy Condition 42 Table ‎4.7 Choice Percentages by Subgroup on Control-Phantom Condition 43 Table ‎4.8 Choice Percentages by Subgroup on Decoy-Phantom Condition 44

(12)

xii

LIST OF FIGURES

Figure ‎2.1 Quasirational Job-Finalist Choice Model (Highhouse, 1997) 21

Figure ‎2.2 Conceptual Framework 23

(13)

1

CHAPTER ONE INTRODUCTION

1.0 INTRODUCTION

This chapter presents the foundation of the present study. It is divided into seven sections: (i) background of the study, (ii) problem statement, (iii) significance of the study, (iv) research objectives, (v) research questions, (vi) definition of terms and (vii) conclusion.

1.1 BACKGROUND OF THE STUDY

Hiring managers are required by organisations to make reliable and valid hiring decisions. However, contextual factors have been found to be able to adversely influence their hiring decisions (Rosati & Stevens, 2009). These contextual factors can lead to a decision bias when hiring managers’ decision-making process is influenced by irrelevant information (Kahneman, Slovic, & Tversky, 1982). During the final stage of personnel selection, hiring managers can be affected by the composition of the job-finalists (i.e. applicants shortlisted for the final stage of selection) and may make mistakes when making their choice (Slaughter, Sinar, & Highhouse, 1999). Job- finalist choice (i.e. personnel selection for the vacant position) is determined by attraction effect, procrastination, over-reliance on detail and representativeness (Highhouse, 1997). Research on the attraction effect, in particular, is more interesting because attraction effect can directly determine which candidate gets chosen due to the composition of candidates. In addition, its research is highly applicable to

(14)

2

industrial and organisational psychology, yet has not received much attention (Dalal et al., 2010).

When assessing the attraction effect on job-finalist choice, decoy and phantom candidates are the terms used to refer to contextual alternatives utilised, and they have been demonstrated to have a powerful effect on job-finalist choice (e.g. Highhouse, 1996; Slaughter et al., 1999; Slaughter, Bagger, & Li, 2006). An attraction effect is a phenomenon where the attractiveness of a target candidate increases relative to a competing candidate when a third alternative (termed decoy candidate or phantom candidate) is introduced, such that the target candidate holds a clear advantage over a decoy candidate while a competing candidate does not, or when a phantom candidate holds a clear advantage over a target candidate while a competitor candidate does not (Sivakumar, 2016). Highhouse and Johnson (1996) explained that target candidate is usually chosen over a competing candidate due to human predisposition to avoid loss rather than to gain benefit.

The attraction effect on job-finalist choice represents a form of bias in the hiring decision when supposedly a target candidate is not the one to be chosen. The effect changes the rank-ordering of candidates based on their composition rather than their attributes, making the selection process unreliable and may lead to bad hire (Highhouse, 1997). Bad hire negatively impacts organisations in terms of low staff morale, drop in productivity, poor client relationships, a decrease in sales, and monetary losses (Maurer, 2015; West, 2013). Such negative consequences of bad hire show the importance of studying attraction effect on job-finalist choice.

(15)

3 1.2 STATEMENT OF PROBLEM

Hiring managers want to make a job-finalist choice free from bias. However, hiring managers’ job-finalist choices can be affected by many decision biases, including the attraction effect (Highhouse, Dalal, & Salas, 2013). Given the negative impact of attraction effect on job-finalist choice, research on the subject ought to be plenteous and well-documented. Research on attraction effect in job-finalist choice has been around for 22 years, since Highhouse (1996). However, research on attraction effect, in general, has been around for 14 years earlier than that (Huber, Payne, & Puto, 1982). This may be the reason why the number of papers published for attraction effect on job-finalist choice remains limited, estimated to be only around six studies (i.e. Connolly, Reb, & Kausel, 2013; Highhouse, 1996; Keck & Tang, 2015;

Slaughter, 2007; Slaughter et al., 2006; Slaughter et al., 1999).

As a result of limited publications, there is a lack of methodological and sample variation shown by previous research. All of the past research used between- subjects design which compared different participants’ job-finalist choices in different groups. This design, however, is vulnerable to the confounding variable of individual differences (Sarason, 1962). Participants’ job-finalist choices might have been confounded by participants’ personal preferences on certain candidates based on additional information of attribute scores provided, rather than the effects of decoy and phantom candidates themselves. For example, instead of being influenced by the attraction effect, participants in control condition might choose one job-finalists because of their preferences on test performance, while participants in experimental condition might choose another job-finalist due to their preferences on interview performance. Therefore, such use of different participants in different conditions might introduce the confounding variable of individual differences. Thus, it is

(16)

4

important to examine whether the attraction effect on job-finalist choice can be replicated using a different experimental design that can remove the confounding effect of individual difference.

In addition, except for Highhouse (1996), all other research only examined decoy-only attraction effect. Researchers post-1996 seemed to be only focusing on the effect of decoy candidate on job-finalist choice due to Highhouse’s finding that the effect of decoy and phantom candidates on job-finalist choice was equal. Compared to job-finalist choice setting, research on phantom in consumer choice setting is still ongoing despite the fact that it was started way before the job-finalist choice setting (e.g. Guney & Richter 2015; Pettibone & Wedell, 2007; Pratkanis & Farquhar, 1992;

Scarpi & Pizzi, 2013; Trueblood & Pettibone, 2015). Not only that, there was a substantial amount of research that incorporated phantom on top of decoy as contextual alternatives of attraction effect in their studies (e.g. Hedgcock, Rao, &

Chen, 2016; Doyle, O’Connor, Reynolds, & Bottomley, 1999; Pettibone, & Wedell, 2000). Highhouse (1996) himself admitted that, while the effect might be equal, they may not be necessarily similar. Both decoy and phantom are considered independent factors of decision biases in judgment and decision making (Highhouse et al., 2013).

The nature of phantom is different from that of decoy. In particular, within job-finalist choice context, a decoy candidate could be an unfavourable candidate in the choice set, while a phantom candidate could be a strong candidate that accepts another job offer or chooses to remain in his or her current position (Highhouse, 1997). Therefore, the present study aimed to include phantom candidate on top of decoy candidate in examining attraction effect on job-finalist choice. Considering the fact that 21 years have passed and one study alone is not enough to be conclusive, it is important to

(17)

5

examine whether the equal effect between decoy and phantom candidates on job- finalist choice can be replicated or not.

In addition, most of the research had used convenient samples consisting of American university students. However, in the real world, job-finalist choice is not made by university students, making their findings hardly generalisable. Highhouse (1996) mentioned the need to examine the effect on experienced hiring decision makers, which has yet to be addressed. It is also important to note that such use of homogenous samples might contribute to the consistent findings and potentially hindered previous research from actually examining attraction effect. Therefore, it is important to examine whether the effect is replicable or not on heterogeneous samples with different backgrounds, especially among experienced hiring decision makers.

In conclusion, there are several problems that aimed to be addressed by the present study. These problems include the frequently used between-subjects design, the lack of comparison between the effects of decoy and phantom candidates, and the overly used convenient samples of American university students.

1.3 SIGNIFICANCE OF THE STUDY

In terms of methodological significance, the present study used within-subjects design where the same participants were administered all conditions. Such design removed the potential confounding individual differences associated with between-subjects design (Sharma, 2017). It is considered to be better than between-subjects design, since it provides the best match in characteristics among participants and has higher statistical power (Charness, Gneezy, & Kuhn, 2012). In addition, the present study used both decoy and phantom candidates to examine the attraction effect. This

(18)

6

allowed the comparison between the effects of the two candidates, and examined whether the similar effect finding by Highhouse (1996) can be replicated or not.

In terms of theoretical significance, the present study was conducted on Malaysian samples. This allowed the present study to investigate whether Malaysian samples’ job-finalist choices can be affected by the presence of decoy and phantom candidates or not. Hofstede (1983) stated that organisational practices, including personnel selection, are dependent upon national culture. This statement was supported by de Mello and Mariano (2004), who found that cultural differences between two countries could explain their variations in recruitment and selection practices. Compared to previous research that used American samples who were individualistic in culture and tend to focus primarily on an object before coming up with a decision, the Malaysian samples used in the present study were collectivistic in culture and tend to view things in a holistic manner before coming up with a decision (Nisbett, Peng, Choi, & Norenzayan, 2001). If the effect is present among the Malaysian samples, the existing theory is strengthened due to its applicability in a different culture. If, on the other hand, it is not, new insights can be gained and would open up a path for future studies to explore.

In addition, the present study used heterogeneous samples consisting of participants with three different backgrounds: (i) personnel selection workers, (ii) personnel selection students and (iii) laypeople. Personnel selection workers were selected to address the need to examine attraction effect on experienced hiring decision makers, as highlighted by Highhouse (1996). Meanwhile, personnel selection students were selected because they were relatively similar to that of university student samples used in previous studies (i.e. the majority were business and psychology majors). Laypeople were selected to serve as a control subgroup since

(19)

7

they did not have knowledge or working experience in personnel selection. Therefore, the use of heterogeneous samples provided an opportunity for the present study to explore possible differences among participants with different backgrounds. This would provide a further understanding on the theory of attraction effect in job-finalist choice and would open up a path for future studies to explore further.

In terms of practical significance, the present study might help in identifying the presence of attraction effect among personnel selection workers. The present study was able to determine whether attraction effect influences personnel selection workers’ job-finalist choices, as the effect did on university students in previous research or not. Such identification is important in efforts towards reducing bias as well as increasing the reliability and validity of job-finalist choice among real hiring decision makers.

In conclusion, the present study has its significance, which includes the use of within-subjects design, the comparison between the effects of decoy and phantom candidates, the use of Malaysian samples and the comparison between participants with different backgrounds, as well as identification on the presence of attraction effect among personnel selection workers.

1.4 RESEARCH OBJECTIVES

This section presents the objectives of the present study. There are three research objectives:

i. To measure the effect of decoy candidate on job-finalist choice ii. To measure the effect of phantom candidate on job-finalist choice

iii. To measure the difference between the effects of decoy and phantom candidates on job-finalist choice

(20)

8 1.5 RESEARCH QUESTIONS

This section presents research questions aimed to be answered in the present study.

There are three research questions:

i. Is there any significant effect of decoy candidate on job-finalist choice?

ii. Is there any significant effect of phantom candidate on job-finalist choice?

iii. Is there any significant difference between the effects of decoy and phantom candidates on job-finalist choice?

1.6 DEFINITION OF TERMS

This section outlines the definition of variables of interest in the present study: (i) attraction effect, (ii) decoy, (iii) phantom and (iv) job-finalist choice.

1.6.1 Attraction Effect

Conceptual definition: Attraction effect is conceptually defined as a situation where the presence of an additional alternative influences the attractiveness of another alternative in the choice set (Keck & Tang, 2015).

Operational definition: In the present study, attraction effect was operationalised as the presence of decoy candidate in decoy condition and phantom candidate in phantom condition presented to participants for hiring decisions. To examine attraction effect, participants were presented with control, decoy and phantom conditions, where their job-finalist choices in each condition were compared.

1.6.2 Decoy

Conceptual definition: Decoy is conceptually defined as an inferior alternative in a choice set that can cause a decision error (Connolly et al., 2013).

(21)

9

Operational definition: In the present study, decoy was operationalised as an additional candidate in the composition of job-finalists in decoy condition presented to participants for hiring decision. This candidate has a similar score in one attribute and a lower score in another attribute when compared to the target candidate. On the other hand, it has a higher score in one attribute and a lower score in another attribute as compared to the competing candidate. To investigate the effect of decoy, participants were presented with control condition (Candidate 1 vs. Candidate 2) and decoy condition (Candidate 1 vs. Candidate 2 vs. Decoy Candidate), and their job-finalist choices in the two conditions were compared.

1.6.3 Phantom

Conceptual definition: Phantom is conceptually defined as a superior alternative in a choice set that is unavailable at the time of the decision, yet still causes a decision error (Pratkanis & Farquhar, 1992).

Operational definition: In the present study, phantom was operationalised as an additional candidate that is unavailable for hire in the composition of job-finalists in phantom condition presented to participants for hiring decision. This candidate has a similar score in one attribute and a higher score in another attribute when compared to the target candidate. On the other hand, it has a higher score in one attribute and a lower score in another attribute when compared to the competing candidate. To examine the effect of phantom, participants were presented with control condition (Candidate 1 vs. Candidate 2) and phantom condition (Candidate 1 vs. Candidate 2 vs.

Phantom Candidate), and their job-finalist choices in the two conditions were compared.

(22)

10 1.6.4 Job-Finalist Choice

Conceptual definition: Job-finalist choice is conceptually defined as a hiring decision made by hiring managers on a group of candidates during the final stage of personnel selection (Highhouse, 1996).

Operational definition: In the present study, job-finalist choice was operationalised as the participants’ choice on one out of two candidates in control and phantom conditions and one out of three candidates in decoy condition.

1.7 CONCLUSION

In this chapter, the following information was presented: research background, problem areas, research importance, research objectives and questions, theoretical and conceptual framework as well as conceptual and operational definitions of variables involved of interest. The next chapter presents a review of previous research relevant to the present study.

(23)

11

CHAPTER TWO LITERATURE REVIEW

2.0 INTRODUCTION

This chapter reviews past research that focused on variables of interest of the present study. This review gives an overview of those variables as well as highlights the gaps existing in past research aimed to be addressed in the present study. This chapter is divided into seven sections: (i) decoy-only attraction effect, (ii) phantom-only attraction effect, (iii) decoy-and-phantom attraction effect, (iv) theoretical framework, (v) conceptual framework, (vi) hypotheses and (vii) conclusion.

2.1 DECOY-ONLY ATTRACTION EFFECT

A classic research by Highhouse (1996) examined the effect of decoy candidate on job-finalist choice. A total of 218 undergraduate psychology students were presented with a simulated personnel selection scenario. Results from logistic regression showed that participants chose the target candidate when there was a decoy candidate, (1, N

= 218) = 24.62, p < .001. This research did not include the control group for baseline data comparison, and this negatively affected its internal validity. Moreover, it used between-subjects design which might suffer from confounding participants’ individual differences. In addition, this research only used American university students with presumably little experience in personnel selection, limiting the generalisability of the findings.

Slaughter et al. (1999) conducted a study examining the issue of generalisability on the effect of decoy candidate in job-finalist choice. A total of 208

(24)

12

introductory psychology students were presented with a 9-minute video vignette of work sample exercises of candidates constructing plastic car models. The candidates in the video produced models with differing quantity and quality, and participants were required to choose one of the candidates. Chi-square analysis indicated that the participants chose target candidate when the decoy candidate was present, (1, N = 173) = 21.45, p < .001. Although this research found significant effect of decoy candidate, Slaughter et al. noted that the effect was not as strong as those found in previous research. Two possible reasons were identified: (i) the influence of different candidate assessment scores from previous research and (ii) the influence of participants having a strong preference for one attribute. This research also did not include phantom candidate. Such exclusion made it not possible for this research to test the effect of phantom candidate and compare it with the effect of decoy candidate.

This research also did not use control group, which lowered its internal validity, since there was no baseline data comparison. It also did not use within-subjects design, which can address the confounding effect of individual difference associated with between-subjects design. In addition, it used convenient samples of American university students as research participants, which limits the generalisability of its findings.

Slaughter et al. (2006) conducted a study to examine the effect of decoy candidate on group-based job-finalist choice. A total of 605 undergraduate business and psychology students were presented with personnel selection scenarios, where they needed to make hiring decisions based on group consensus. Results showed that participants chose target candidate when there was a decoy candidate, (2, N = 111)

= 18.50, p < .01. This research did use a control group, which increased its internal validity through baseline data comparison. However, it did not include a phantom