DECISION MAKING UNDER TIME SCARCITY

(1)

DECISION MAKING

UNDER TIME SCARCITY

(2)

DECISION MAKING

UNDER TIME SCARCITY

The effect of time scarcity on decision making

performance under individual variation in action

identification

By Shirin Mantik

16

th

June 2016

to

1

st

_{Supervisor: Prof. Dr. Bob Fennis}

2

nd

Supervisor: Dr. Yannick Joye

Faculteit Economie en Bedrijfskunde Nettelbosje 2

9747 AE Groningen, Nederland

University of Groningen

MSc Marketing Management Department of Marketing Faculty of Economics and Business

(3)

I

ABSTRACT

Time is our most limited and therefore one of our most important resources. It passes constantly and especially in our daily working environment we have to manage it carefully in order to make the most efficient decisions. But what effect does a sudden lack of time have on our decision making process? This study provides further evidence that the time available to make decisions definitely has an impact on the choices we make and on the quality of our final decisions. The Iowa Gambling Task, was putting participants in a self-control dilemma by letting them choose between smaller immediate rewards and larger rewards in the long run. It revealed that time pressure increased risk affinity, impatience and led to an overall lower gambling value of participants under time pressure. Additionally, decision making performance was measured with the apartment task by Dijksterhuis (2004), where participants had to evaluate four different apartments. The apartment task focused on the ability to quickly process a large amount of information and revealed that time scarcity limited the amount of information that could be memorized. Thus, it reduces our probability to estimate the best option. Moreover, this study accounted for individual variation in action identification, which was measured by the Behavioral Identification Form (BIF). Against expectations that a low BIF level would lead to worse decision performances due to a higher susceptibility to environmental cues, the study revealed that the BIF level had no influence on the IGT performance and only a marginal negative impact of a low BIF level on the apartment task. Hence, the effect of individual action identification was shown to be depending on the type of decision making task. A low BIF level was expected to strengthen the negative effect of time scarcity on decision performances, since low level agents were expected to be more distracted by time constraints. However, high level agents performed significantly worse in the IGT under time scarcity while time pressure had no remarkable influence on the performance of low level agents. Thus, a high BIF level actually boosts the negative effect of time scarcity and the study could proof that individuals with a rather mechanic-tendency to think, like low level agents, can indeed counteract the negative effect of time scarcity. A strong focus on the current action helps thereby to approach rather intuitive decision making tasks, like the IGT under time pressure. In the end we do not are subordinates to our pre-existing personality or to external conditions but still architects of our own decision making performance.

(4)

II

LIST OF FIGURES

Figure 1: Model of decision making under time stress, Zakay 1993 ...- 3 -

Figure 2: Expected interaction effect ...- 6 -

Figure 3: Conceptual Model ...- 6 -

Figure 4: Study Design ...- 7 -

Figure 5: IGT Introduction Text ...- 7 -

Figure 6: Apartment Choice Task Introduction Text ...- 8 -

Figure 7: BIF Introduction Text ...- 8 -

Figure 8: Choosing from deck B with a reward of $100 ...- 11 -

Figure 9: Attribute “Landlord” for Time Scarcity Condition ...- 12 -

Figure 10: Effect of condition of participants on risky and safe decisions in IGT ...- 14 -

Figure 11: Interaction Effect on Risky Decisions ...- 15 -

Figure 12: Interaction Effect on Safe Decisions ...- 16 -

(7)

V

LIST OF ABBREVIATIONS

BIF……. Behavioral Identification Form DV……. Dependent Variable (statistical) IGT……. Iowa Gambling Task

IV……. ..Independent Variable (statistical) M…….... Mean (statistical)

SD……...Standard Deviation (statistical) TS……...Time Scarcity

(8)

- 1 -

INTRODUCTION

We all know the pressuring feeling of deadlines coming closer, to-do lists getting longer leading to meetings that have to be rescheduled. The scarcity humans experience is no longer restricted to poverty or the urgent feeling of being hungry. The scarcity we experience in the 21st century and which becomes more and more relevant, is scarcity of time. Having less than you feel you need is scarcity per definition according to Mullainathan (2013), and results in a feeling we all know in form of tension or stress. Time is our most limited and therefore one of our most important resources. It passes constantly and especially in our daily working environment we have to manage it carefully in order to make the most efficient decisions. But what effect does a sudden lack of time have on our decision making process?

Since decision making is a complex cognitive activity, which is very sensitive to situational and environmental conditions (Payne, 1982), there is no doubt that a shortage of time clearly has an effect on our final decision making quality (e.g. Klapproth, 2008; Payne, 1982). Yet, results of existing empirical research show a non-uniform effect of time scarcity, depending on different types of decisions and external circumstances. Although some studies outline a positive relation (Dambacher & Hübner, 2015; Maule, Hockey, & Bodzola, 2000), due to faster or more effortful information processing under time pressure, most authors emphasize the negative effect of time pressure on decision making (Diederich, 2003; Pachella & Pew, 1968; Ruthruff, 1996; Zakay, 1993), which is discussed further in the research framework.

This study aims at a deeper examination of the relation between time scarcity and decision making performance. It was conducted as a field experiment under actual induced time scarcity, in order to reflect everyday decision making conditions as close to reality as possible, since previous research investigated the effect of time scarcity mainly in a laboratory setting. Moreover, the experiment included two different kinds of decision making tasks in order to enable comparability of effects across results.

(9)

- 2 - action identification is measured with the Behavioral Identification Form (BIF) by Vallacher & Wegner (1989). Hence, the problem statement this study investigates is derived as:

What effect does time scarcity have on decision making performance and how is this effect influenced by individual variation in action identification, measured by the BIF?

Subsequently, the research frame addresses each part of the problem statement by giving a definition of the terms scarcity, decision making performance and variation in action identification. Further, assumed relations are deduced (H1, H2 and H3), based on existing literature. The methodology outlines the procedure of the experiment as well measurement methods as a basis for the analysis. Accordingly, results of the experiment are discussed, leading to an overall conclusion of the study.

RESEARCH FRAMEWORK

Definition of TIME SCARCITY

(10)

- 3 -

Figure 1: Model of decision making under time stress by Zakay (1993)

Zakay (1993) depicts that objective time constraints as well as subjective perception of time both lead to an increased awareness towards the passage of time and a prospective estimation of time. This results in a division of attentional resources for information processing.

(11)

- 4 - Additionally, a shortage of cognitive resources leads to performance errors, which results with simplified strategies, in suboptimal decision making performances.

Definition of DECISION MAKING PERFORMANCE

Decision making can be defined as the evaluation of at least two alternatives, differing on a number of attributes, whereas the selection of an alternative depends especially on the actual time available to make that decision (Ariely & Zakay, 2001). Decision making was first measured with the Iowa Gambling Task (IGT) by Bechara et al. (1994) and used to assess the performance on a complex decision task with uncertain outcomes. Participants had to make a number of selections via a card game. They had to choose between an immediate but smaller reward or a larger reward in the long-run putting them into a self-control dilemma (De Ridder et al., 2014). Decision making was further tested with the Apartment Choice Task by Dijksterhuis (2004), which was referring to the very basic definition of decision making as choosing between alternatives differing on a number of attributes. Participants received four different descriptions of apartments and had to evaluate their attractiveness based on the valence of attributes they were exposed to beforehand. While the IGT addressed gambling behavior, the apartment task focused on the ability to (quickly) process a large amount of information.

Expected Results main effect time scarcity

.

Based on the previously introduced decision making model by Zakay (1993), a suboptimal decision making performance from participants with a time limit is expected. Relating this to the Iowa Gambling Task, a study by Shah (2014) supports the expected outcome by stating that participants under time scarcity are stressed and therefore less patient, which results in steeper discounting of future values. This leads to the assumption that participants under time scarcity will rather choose from card decks yielding a high immediate reward, but that will result in a lower long-term value. Implications for the apartment task are based on a different effect of time scarcity but are expected to result as well in a lower performance level: while the IGT challenges patience in self-control dilemmas, the apartment task focuses on the ability to process and memorize an overload of information. The unconscious thought theory (UTT) by Dijksterhuis and Nordgren (2006) showed that immediate thinkers approached the apartment task worst: only 36% could detect the best apartment. Since immediate thinkers had no time for a proper evaluation of the four alternatives, the same negative effect is expected to occur for participants under time pressure in this study. Both expectations are summarized in the first hypothesis:

(12)

- 5 -

Definition of INDIVIDUAL VARIATION IN ACTION IDENTIFICATION

In order to account for differences in personality traits, the individual variation in action identification was measured by the Behavioral Identification Form (BIF) by Vallacher and Wegner (1989). Based on this instrument, actions are distinguished in “Low level identities”, specifying how an action is performed and “High level identities”, specifying why something is performed. Thus, individual differences in action identification measure the broad personality of a person: high level agents view their actions in terms of causal effects and personal goals. They generally process actions with abstract features in mind (Vallacher & Wegner, 1989). In contrast, low level agents tend to think about their actions in detail at a more uniform level, based on procedural aspects (Belayachi & Van der Linden, 2012).

Expected Results main effect BIF level. While high level agents are characterized by a good ability to adopt an appropriate level of identification for different actions, low level agents have more difficulties in understanding unknown actions (Belayachi & Van der Linden, 2012). In line with that, Vallacher and Wegner (1989) state that low level agents report a greater tendency to make action errors. Referring this to the model of Zakay (1993), low level agents are expected to perform worse, since cognitive errors lead to suboptimal performances. However, it is unclear whether low level agents perform equally worse in the IGT and the apartment task or if any differences in performances occur. Generally, it is assumed that:

H2

:

A low BIF has a negative effect on decision making performance

Expected Results interaction effect. Since a low BIF level causes higher susceptibility to environmental cues due to a lack of inner motivational drivers, low level agents are expected to be more susceptible to time scarcity than high level agents (cf. figure 2). Therefore, participants with a low BIF will be stronger distracted by a shortage of time than participant with a high BIF, leading to a significant worse decision making performance:

H3

:

A low BIF is strengthening the negative effect of time scarcity on decision making

(13)

- 6 -

Figure 2: Expected interaction effect of BIF level and Time Scarcity

The expected results (H1-H3) are displayed in the following conceptual model:

Figure 3: Conceptual Model

METHODOLOGY

PARTICIPANTS AND DESIGN

(14)

- 7 -

Figure 4: Study Design

EXPERIMENT PROCEDURE

The participants of this study were approached at the various faculties of the University of Groningen. The experimenter placed a table with a tablet, a laptop and cookies as incentives at the university canteen and libraries. The voluntary participants were told that they were about to join a quantitative research experiment with different tasks, concerning a Master thesis study about decision making and that they will receive cookies for joining. They were randomly assigned to either the TS-condition (50 participants) or the neutral condition (50 participants). First participants were asked to approach the Iowa Gambling Task, which was presented as an App on a tablet. The IGT-App was downloaded for free in the Android App-Store (cf. Appendix A). After the experimenter entered the number of the participant (cf. Appendix A), to relate the incoming results to other measurements, the participants were given the tablet and got instructed to read the following introduction text:

(15)

- 8 - Participants in the neutral condition were asked to start the task by clicking “Continue” (cf. Appendix A). Participants in the TS-condition were informed that they had two minutes to maximize their fictive starting value of $2.000. Afterwards, the experimenter placed a mobile phone with a 2-minute countdown on the desk in front of the TS-condition participants to indicate the remaining time. After finishing the IGT, participants received the Qualtrics survey on the tablet and read the introduction text for the apartment task:

Figure 6: Apartment Choice Task Introduction Text

For participants in the TS-condition the additional information was provided: “Each attribute will be shown for 3 seconds. Please choose immediately afterwards.” Then, both groups saw twelve apartment attributes in sequence and were requested to evaluate the apartments afterwards (cf. Appendix B). After rating the apartments, both groups filled out the BIF accompanied by the explanation:

Figure 7: BIF Introduction Text

(16)

- 9 -

MANIPULATION & MEASUREMENT OF INDEPENDENT VARIABLES

Time Scarcity Manipulation. According to MacGregor (1993), a common method to induce stress is the limitation of available time for decision making. Thus, in the TS-condition, decisions must be made within certain deadlines (Klapproth, 2008). In order to manipulate time scarcity, a time optimum (average time needed to fulfill a task without time pressure) had to be found (Mac Gregor, 1993). The average time needed to finish the IGT, was measured through trial and error and accounted approximately for three to three and a half minutes. It was reduced to two minutes in order to induce a sufficient level of time pressure. The actual time countdown of two minutes was shown on a mobile phone clock in front of the participants in the TS-condition. Additionally, they were informed that they had to finish the IGT within the two minutes. However, if participants could not finish within the given time frame the experiment was stopped and participants were excluded from the further procedure. According to the apartment task, an exposure time of three seconds (Dijksterhuis, 2004) was set for each attribute (with the advanced question type “Timing” in Qualtrics) before the next attribute appeared automatically (cf. figure 9). In contrast, participants in the neutral condition had to click on the “next” button to see the next attribute, which allowed them to evaluate attributes as long as they wanted. Limiting time for both tasks was expected to change choices of the participants as there was less time to process the given information and less time to compare all alternatives. Due to less accuracy and thereby more cognitive errors, poorer final decision performances were assumed for participants in the TS-condition. This supports the first hypothesis. At the end of the survey, the manipulation of time scarcity was tested by the direct question: “How do you feel towards the experiment?” and the statement “I felt time pressure during the experiment” (Edland, 1993). Participants could agree on a Likert-scale (0 = I do not agree at all, 7 = I absolutely agree, cf. Appendix D). Participants in the TS-condition were expected to report greater feelings of time pressure.

Behavioral Identification Form Measurement. In order to measure the level of individual action identification, the BIF was applied. Each item of the BIF represents an act identity followed by one lower alternative (0), which reflects how an action is performed as well as one higher alternative (1), specifying rather why an action is performed (cf. Appendix C). Participants were asked to choose the alternative that described the action best for them, for instance:

(17)

- 10 - o sitting in a chair (0)

o looking at the blackboard (1)

The level of individual action identification was measured as the number of high level alternatives (1) ticked on the BIF, implying that participants who chose more high level alternatives than low level alternatives, score high on the BIF and can be defined as “high level agents”. In order to categorize participants into either high or low level agents a median split was used: the twenty-five BIF items were summarized for each participant and then split at an assessed median of 15. This assigned 55,6% of participants with a mean BIF of 15 and higher a one (1=high BIF) and participants with a BIF mean lower than 15 a zero (0=low BIF), resulting in the binary variable BIFlevel, which was used for the further analysis as the measured independent variable. Low level agents which are more susceptible to external and situational cues were expected to perform poorer in both decision making tasks (Vallacher & Wegner, 1998). Therefore, measurements of a low BIF level were expected to support the second hypothesis, by reflecting worse decision performances.

MEASUREMENT OF DEPENDENT VARIABLES

(18)

- 11 -

Figure 8: Choosing from deck B with a reward of $100

(19)

- 12 -

Apartment Choice Task Measurement. The apartment choice task by Dijksterhuis (2004) focuses on the ability to quickly process a large amount of information and to correctly detect the best option among alternatives. Participants received four descriptions of apartments, varying on twelve attributes (e.g. area, price, size). All apartments were described by positive and negative attributes. Apartment B was described rather desirable (8 positive, 4 negative attributes), apartment D rather negative (8 negative, 4 positive attributes) and the other two apartments were described neutral in value (6 positive, 6 negative attributes) and served thereby as fillers in order to increase task complexity. The attributes were presented after each other and for all of the four apartments at the same time (cf. Appendix B). In the TS-condition each attribute was shown for three seconds (presented as a countdown above the attribute) before the next attribute appeared automatically, as displayed by figure 9:

Figure 9: Attribute “Landlord” for Time Scarcity Condition

(20)

- 13 - (Edland & Svenson, 1993). This presumably leads to a suboptimal apartment evaluation, supporting the second hypothesis.

RESULTS

RANDOMIZATION AND MANIPULATION CHECK

Randomization check. Multiple Analyses of Variances (ANOVAs) with time scarcity and BIF

level as independent variables and age and gender as dependent variables, proved the absence of any effects on gender due to time scarcity or the BIF (p > .05). A significant effect of time scarcity on age (F(3,96) = 11.18, p = .001) could be found. However, the analysis revealed only a small difference in age of participants in the time scarcity condition (M = 21.97, SD = 0.373) and participants in the neutral condition (M = 23.71, SD = 0.363), which is not assumed to influence any decision outcomes.

Manipulation check. In order to test the validity of the time scarcity manipulation, an ANOVA with TS-condition as independent variable and the reported feeling of time pressure during the experiment as the dependent variable yielded a highly significant effect (F(3,96) = 27.187, p = .000). As expected participants in the TS-condition reported significantly higher feelings of time pressure during the experiment (M = 3.98, SD = 1.83) than participants in the neutral condition (M = 2.26, SD = 1.44, cf. Appendix E). Moreover, participants in the TS-condition reported a significant difference in mood after the experiment (F(3,96) = 9.817, p = .002) compared to participants in the neutral condition (M = 5.84, SD = .888). The more negative rating of mood from participants under time pressure (M = 5.24, SD = 1.021) can be explained due to stress caused by the time constraint and further verifies a valid manipulation.

IOWA GAMBLING TASK RESULTS

(21)

- 14 - from the safe, “advantageous” decks C and D (M = 57.08, SD = 13.73) than participants in the TS-condition (M = 50.04, SD = 14.69). Both results are displayed by figure 10:

Figure 10: Effect of condition of participants on risky and safe decisions in IGT

It is also pictured by figure 10 that participants in the TS-condition choose on average 50% from the safe decks and 50% from the risky decks and thus show no obvious “bad decision” tendency, in terms of a higher number of risky choices. Only when compared to participants in the neutral condition, their performance can be defined as poorer.

Additionally the analysis revealed that participants under time pressure marginal significantly (F(1,96) = 3.357, p = .070) gained a smaller net value in the IGT ($1.641) than participants in the control condition ($1.936) which further supports the first hypothesis that participants with time scarcity performed worse on the IGT (cf. Appendix F).

(22)

- 15 - the overall net value yielded similar non-significant results (p > .01). Hence, the second hypothesis that a negative effect of the BIF might occur, could not be significantly validated.

Interaction Effect. Regarding the interaction effect, the full factorial analysis also tested whether a low BIF strengthens the negative effect of time scarcity on the IGT performance, as assumed by the third hypothesis. Results reveal a significant interaction effect on risky decisions (F(1,96) = 4.432, p = .038), as well as safe decisions (F(1,96) = 4.768, p = .031), which are represented in figures 11 and 12. Low level agents in the neutral condition choose more often from risky decks (M = 47.37, SD = 14.06) than high level agents (M = 39.54, SD = 11.81). Surprising is that low level agents made less risky decisions under time scarcity (M = 47.70, SD = 18.32) than high level agents (M = 51.76, SD = 12.26). As displayed by figure 11, the performance of low level agents did not remarkably vary across conditions, whereas high level agents performed significantly worse under time scarcity than in the neutral condition. Against expectations, low level agents could be less distracted by the time countdown due to a strong focus on the current action, so that their division of attentional resources towards the time constraint was weaker. The reversed effects occurred for safe decisions (cf. figure 12). Both results do not validate the third hypothesis, that a low BIF strengthens the negative effect of time scarcity but an opposite interaction effect: a high BIF significantly strengthens the negative effect of time scarcity on the IGT performance, whereas a low BIF rather seems to weaken it.

(23)

- 16 -

Figure 12: Interaction Effect on Safe Decisions

APARTMENT TASK

Main Effect of Time Scarcity. Decision making was also measured with the Apartment Task.

(24)

- 17 - SD = 2.59). The apartment task therefore supports the first hypothesis that time scarcity has a negative effect on decision making performance:

Figure 13: Effect of participant condition on the difference in attitude towards apartment B and D

Main Effect of BIF Level. Additionally, the analysis revealed that participants with a low BIF performed marginal significantly (F(1,96) = 3.48, p = .065) worse in the apartment task (M = 1.68, SD = 2.55) than participants with a high BIF (M = 2.46, SD = 2.56), since their difference score was lower (cf. Appendix G). Hence, the second hypothesis that a low BIF has a negative impact on decision making could be confirmed as displayed by figure 14:

(25)

- 18 -

Interaction Effect. The full-factorial analysis yielded a non-significant effect (F(1,96) = 0.004, p = .950) for interaction effect of time scarcity and the BIF on the difference score of apartment ratings. Thus, the third hypothesis, that a low BIF is significantly strengthening the negative effect of time scarcity on decision making could not be supported by the apartment task. Results for the IGT and Apartment task are consolidated in figure 15:

Figure 15: Conceptual Model Results

DISCUSSION

BAD NEWS: WE PERFORM WORSE UNDER TIME PRESSURE

The study sought to provide an answer to the question whether we perform indeed worse under time pressure, since existing empirical research yielded non-uniform results. The rather unpleasant answer that this study ascertained is: yes.

(26)

- 19 - According to the apartment task, participants under time scarcity performed worse since their difference in apartment ratings between the best and the worst alternative was smaller. This implicated that participants with a time limit were not as content about the “best” or “worst” alternative and rather choose attractiveness ratings close to each other. The increased level of psychological stress might have impacted their cognitive working memory, which was necessary in this task to memorize the given information (Kahneman, 1973). Another explanation besides the lacking possibility to memorize all attributes, could also be premature closure (Janis, 1983), meaning that participants who had the impression that the task was not “manageable” in the given time frame, could just have stopped to pay attention and guessed randomly when it came to the rating. This result supports the Unconscious thought theory by Dijksterhuis (2004), which depicts worse performances from “immediate thinkers”.

GOOD NEWS: PERSONALITY NOT AS SHAPING AS EXPECTED

Findings for the effect of the BIF level were not that distinct. Against expectations no significant effect was found for the IGT, whereas a low BIF level had a marginal impact on the apartment task performance. This disparity in results is likely to be caused due to the difference in both types of decision making tasks. The request of the IGT was not straightforward (“choose one card after another and see what is under each card”) and the outcome thereby quite uncertain. All that was required was to mechanically choose one card after another. Since a low action identification level represents the tendency to see an action in terms of its details or mechanics (Vallacher & Wegner, 1989) some low level agents probably performed better in the IGT, since they were not questioning any causal effects. This mechanic-thinking tendency of some low level agents probably led to an overall non-significant effect of the BIF level on the IGT. In contrast, the apartment task was rather goal focused (“please evaluate the four different apartments according to different attributes”) and thus easier to approach for high-level agents, which have less difficulties in understanding actions with goal representations. Taking both results into conclusion, low level agents still seemed to have in general more problems in approaching decision tasks, especially when they are goal focused. However, this effect holds not vice versa: tasks that are less goal focused do not lead to significantly better performances of low level agents.

(27)

- 20 - DON’T LET TIME PRESSURE DISCOMPOSE YOU!

Although, the BIF level had no main influence on the IGT, it significantly moderated the effect of time scarcity. Low level agents were expected to be more distracted by time scarcity, due to their greater focus on external cues (cf. figure 2). However, the reversed effect occurred. It was only of little importance for low level agents whether there was a time limit or not: the amount of safe decisions stayed almost the same (cf. figure 11 & 12). Against expectations the factor of time scarcity played a major role for high level agents. The analysis revealed the striking results that high level agents performed even worse than low level agents under time scarcity but way better in comparison without any time constraint. This results indicate that very intuitive tasks (like the IGT) do not require inner motivational drivers or higher goals under time pressure but rather a focus on the current action and pattern. Counterintuitively, a focus on the current situation does not deteriorate decision making performance under time scarcity. This might be caused due to efficiently blanking out the time pressure, so no division of attentional resources occurred. However, thinking about the outcomes and possible goals of an action resulted in better decision making performances under the condition that there was enough time to come to a conclusion. High level agents who had an unlimited amount of time performed better on the IGT than low level agents because they were able to indicate and differentiate the two safe decks from the two risky ones.

LIMITATIONS

A more regular distribution among the BIF level could improve the significance of results: since there were more high level agents (M = 56) than low level agents (M = 44, cf. Appendix E), the overall decision making performance might have been quite high, which lead to rather good decisions. Including more low level agents (by using a bigger sample) could hold more diversified results and could improve the significance level of findings for the IGT. Since a low BIF showed a negative trend towards decision making performance, a broader research with more participants and various decision making tasks, might also reveal an interaction effect between the BIF and time scarcity on the apartment task.

CONCLUSION AND IMPLICATIONS

(28)

(29)

- 22 -

REFERENCES

Ariely, D. & Zakay, D. (2001). A timely account of the role of duration in decision making. Acta Psychologica, 108, 187-207.

Bechara et al. (1994). Insensitivity to future consequences following damage to human pre-frontal cortex. Cognition, 50, 7-15.

Ben Zur, H. & Breznitz, S.J. (1981). The effect of time pressure on risky choice behaviour. Acta Psychologica, 47, 89-104.

Belayachi, S. & Van der Linden, M. (2012). Individual differences in cognitive representations of action influence the activation of goal concepts. Acta Psychologica, 142, 259-264.

Dambacher, M. & Hübner, R. (2015). Time Pressure affects the efficiency of perceptual processing in decisions under conflict. Psychological Research, 79, 83-94.

De Ridder et al. (2014). Always Gamble on an Empty Stomach: Hunger associated with advantageous decision making. Plosone, 9(10), 1-8.

Diederich, A. (2003). MDFT account of decision making under time pressure. Psychonomic Bulleting & Review, 10, 157-166.

Dijksterhuis, A. (2004). Think Different: The Merits of Unconscious Thought in Preference Development and Decision Making. Journal of Personality and Social Psychology, 87(5), 586-598.

Dijksteruis, A. & Nordgren, L.F. (2006). A Theory of Unconscious Thought. Perspectives on Psychological Science, 1(2), 95-109.

Edland, A. (1993). The Effects of Time Pressure on Choices and Judgements of Candidates to a University Program. In Svenson, O. & Maule, A.J. (Eds), Time Pressure and Stress in Human Judgement and Decision Making (pp. 73-82). New York: Plenum Press.

Edland, A. & Svenson, O. (1993). Judgement and Decision Making Under Time Pressure: Studies and Findings. In Svenson, O. & Maule, A.J. (Eds), Time Pressure and Stress in Human Judgement and Decision Making (pp. 73-82). New York: Plenum Press.

(30)

- 23 - Klapproth, F. (2008). Time and decision making in humans. Cognitive, Affective, & Behavioral Neuroscience, 8(4), 509-524.

MacGregor, D. (1993). Time Pressure and Task Adaption. In Svenson, O. & Maule, A.J. (Eds), Time Pressure and Stress in Human Judgement and Decision Making (pp. 73-82). New York: Plenum Press.

Maule, A.J., Hockey, G.R.J. & Bdzola, L. (2000). Effects of time pressure on decision-making under uncertainty: Changes in affective state and information processing strategy. Acta Psychologica, 104, 283-301.

Miller, J.G. (1960). Information input overload and psychopathology. American Journal of Psychiatry, 116, 695-704.

Mullainathan, S. & Shafir, E. (2013): Scarcity- the new science of having less and how it defines our lives. New York: Picador.

Pachella, R.G. & Pew, R.W. (1968). Speed-accuracy trade-off in reaction time: Effect of discrete criterion times. Journal of Experimental Psychology, 76, 19-24.

Payne, J.W. (1982). Contingent and Decision Behavior. Psychological Bulletin, 93, 382-402. Payne, J.W., Bettman, J.R. & Johnson, E.J. (1993). The use of multiple strategies in judgement and choice. In N.J. Castellan (Ed.), Individual and group decision making (19-39). Hillsdale, NJ: Erlbaum.

Ruthruff, E. (1996). A test of the deadline model for speed-accuracy trade-offs. Perception & Psychophysics, 57, 56-64.

Shah, A.K. (2014). Social Class and Scarcity: Understanding consumers who have less. In Norton, M.I.; Rucker, D.D. & Lamberton, C. (Eds), The Cambridge Handbook of Consumer Psychology (pp. 673-692). Hovedbiblioteket.

Simon, H.A. (1955). A behavioral model of rational choice. Quarterly Journal of Economics, 69, 99-118.

Svenson, O. & Maule, A.J. (Eds), Time Pressure and Stress in Human Judgement and Decision Making. New York: Plenum Press.

(31)

- 24 - Wilson, T.D. & Schooler, J.W. (1991). Thinking too much: Introspection can reduce the quality of preferences and decisions. Journal of Personality and Social Psychology, 60, 181-192. Wright, P. (1974). The harassed decision maker: Time pressures, distractions, and the use of evidence. Journal of Applied Psychology, 59(5), 555-561.

Zakay, D. (1993). The Impact of Time Perception Processes on Decision Making under Time Stress. In Svenson, O. & Maule, A.J. (Eds), Time Pressure and Stress in Human Judgement and Decision Making (pp. 59-72). New York: Plenum Press.

(32)

- 25 -

APPENDICES

Appendix A

(33)

(34)

- 27 -

Appendix B

Apartment Choice Task for participants in time scarcity condition from Qualtrics

(35)

- 28 -

Q2 Bus Stop

Q3 Landlord

(36)

- 29 -

Q5 Bathroom

Q6 Size

(37)

- 30 -

Q8 Balcony

Q9 Area

(38)

- 31 -

Q11 Pets

(39)

- 32 -

Appendix C

Behavioral Identification Form by Vallacher and Wegner (1989) from Qualtrics

We are now interested in your personal preferences in different behavior descriptions. Therefore, you will receive a number of actions and each time two options (A and B) describing that action. Please tick the option that describes the action best in your opinion. Do not hesitate with your answer but choose the option that first comes to your mind.

1.Making a list

Getting organized (1) Writing things down (0)

2.Reading

Following lines of print (0) Gaining knowledge (1)

3.Joining the Army

Helping the Nation's defense (1) Signing up (0)

4.Washing Clothes

Removing odors from clothes (1) Putting clothes into the machine (0)

5.Picking an apple

Getting something to eat (1) Pulling an apple off a branch (0)

6.Chopping down a tree

Wielding an axe (0) Getting firewood (1)

7. Measuring a room for carpeting

Getting ready to remodel (1) Using a yardstick (0)

8. Cleaning the house

(40)

- 33 -

9. Painting a room

Applying brush strokes (0) Making the room look fresh (1)

10. Paying the rent

Maintaining a place to live (0) Writing a check (1)

11. Caring for houseplants

Watering plants (0) Making the room look nice (1)

12. Locking the door

Putting a key in the lock (0) Securing the house (1)

13. Voting

Influencing the election (1) Marking a ballot (0)

14. Climbing a tree

Getting a good view (1) Holding on to branches (0)

15. Filling out a personality test

Answering questions (0) Revealing what you're like (1)

16. Toothbrushing

Preventing tooth decay (1) Moving a brush around in one's mouth (0)

17. Taking a test

(41)

- 34 -

18. Greeting someone

Saying hello (0) Showing friendliness (1)

19. Resisting temptation

Saying "no" (0) Showing moral courage (1)

20. Eating

Getting nutrition (1) Chewing and swallowing (0)

21. Growing a garden

Planting seeds (0) Getting fresh vegetables (1)

22. Traveling by car

Following a map (0) Seeing the countryside (1)

23. Having a cavity filled

Protecting your teeth (1) Going to the dentist (0)

24. Talking to a child

Teaching a child something (1) Using simple words (0)

25. Pushing a doorbell

Moving a finger (0) Seeing if someone's home (1)

(42)

- 35 -

Appendix D

(43)

- 36 -

Appendix E

Randomization, Manipulation and Mood Check

1) Randomization Check: ANOVA with Age, Gender and Mood as independent variables and Time Scarcity and BIF level as dependent variables

Between-Subjects Design

Label N

BIFlevel ,00 low BIF 42

1,00 high BIF 53

Prime ,00 neutral 47

1,00 Time Scarcity 48

Descriptives Abhängige Variable: Age

BIFlevel Prime Mittelwert

Standardabweic

hung N

low BIF neutral 24,0455 2,53504 22

TS 21,5500 2,41650 20

Gesamt 22,8571 2,75479 42

high BIF neutral 23,4400 2,31084 25

TS 22,3929 2,54354 28

Gesamt 22,8868 2,47029 53

Gesamt neutral 23,7234 2,41104 47

TS 22,0417 2,50071 48

Gesamt 22,8737 2,58573 95

2) Manipulation Check: ANOVA with Time Scarcity as IV and “How do you feel towards the experiment”-rating as DV

Factor 1: Time Scarcity Factor2: BIF

(44)

- 37 -

N Mean Std. Deviation Std. Error

neutral 47 2,3404 1,44879 ,21133 Time Scarcity 48 3,9375 1,86121 ,26864 Total 95 3,1474 1,84488 ,18928 SS df MS F Sig. Between Groups 60,571 1 60,571 21,719 ,000 Within Groups 259,366 93 2,789 Total 319,937 94

3) Mood Check: ANOVA with time scarcity as IV and “How do you feel right now"- Rating as DV

Quelle df Mean Square F Sig.

Adjusted Modell 3 4,450 4,915 ,003 Constant Term 1 2815,146 3109,067 ,000 BIFlevel 1 4,235 4,677 ,033 Prime 1 8,533 9,424 ,003 BIFlevel * Prime 1 ,556 ,614 ,435 Error 91 ,905 Total 95 Korrigierte Gesamtvariation 94

Appendix F

Reliability Analysis of the factor BIFlevel

1) Univariate Analysis with Time Scarcity and BIF level as IV and Risky Decisions, Safe Decisions and Net value as DV

Cronbachs Alpha No. Items

,802 25

2) Univariate Analysis with Time Scarcity and BIF level as IV and Risky Decisions as DV

BIFlevel Prime N

low BIF neutral 22

Time Scarcity 20

Total 42

(45)

- 38 - Time Scarcity 28 Total 53 Total neutral 47 Time Scarcity 48 Total 95

Adj. Model 3 651,179 3,209 ,027 Constant Term 1 205327,584 1011,735 ,000 BIFlevel 1 107,611 ,530 ,468 Prime 1 763,009 3,760 ,056 BIFlevel * Prime 1 897,109 4,420 ,038 Error 91 202,946 Total 95 Korrigierte Gesamtvariation 94

3) Univariate Analysis with Time Scarcity and BIF level as IV and Safe Decisions as DV 4)

Descriptives Abhängige Variable: Safe Decisions

BIFlevel Prime Mittelwert

Standardabweic

hung N

low BIF neutral 51,8182 14,39126 22

TS 52,3000 18,32485 20

Gesamt 52,0476 16,17881 42

high BIF neutral 60,9200 12,46635 25

TS 48,5714 12,14550 28 Gesamt 54,3962 13,67668 53 Gesamt neutral 56,6596 14,02525 47 TS 50,1250 14,96467 48 Gesamt 53,3579 14,79931 95 Quelle Quadratsumme

vom Typ III df

(46)

- 39 -

a. R-Quadrat = ,104 (korrigiertes R-Quadrat = ,075)

5) Univariate Analysis with Time Scarcity and BIF level as IV and Net Value as DV

Descriptives Abhängige Variable: Net Value

BIFlevel Prime Mean Std. Deviation N

low BIF Neutral 1747,7273 684,56230 22

Time Scarcity 1747,5000 886,59238 20

Total 1747,6190 777,36283 42

high BIF Neutral 2077,0000 681,33295 25

Time Scarcity 1548,2143 735,25407 28

Total 1797,6415 752,30837 53

total Neutral 1922,8723 695,49853 47

Time Scarcity 1631,2500 798,86056 48

Total 1775,5263 759,79402 95

Adj. Model 3 1250550,535 2,253 ,088 Constant Term 1 296202805,400 533,611 ,000 BIFlevel 1 98713,203 ,178 ,674 Prime 1 1634963,203 2,945 ,090 BIFlevel * Prime 1 1632154,781 2,940 ,090 Error 91 555091,451 Total 95 Korrigierte Gesamtvariation 94

Appendix G

(47)

- 40 -

Descriptives

BIFlevel Prime Mean Std. Deviation N

low BIF neutral 2,5455 2,53973 22

Time Scarcity ,7500 2,40340 20

Total 1,6905 2,60850 42

high BIF neutral 3,3600 2,17715 25

Time Scarcity 1,8214 2,72238 28

Total 2,5472 2,57627 53

Total neutral 2,9787 2,36358 47

Time Scarcity 1,3750 2,62253 48

Total 2,1684 2,61197 95

DECISION MAKING UNDER TIME SCARCITY

DECISION MAKING

UNDER TIME SCARCITY

DECISION MAKING

UNDER TIME SCARCITY

The effect of time scarcity on decision making

performance under individual variation in action

identification

By Shirin Mantik

16

June 2016

to

1

Supervisor: Prof. Dr. Bob Fennis

2

Supervisor: Dr. Yannick Joye

University of Groningen

ABSTRACT

TABLE OF CONTENTS

LIST OF FIGURES

LIST OF ABBREVIATIONS

INTRODUCTION

RESEARCH FRAMEWORK

Definition of TIME SCARCITY

Definition of DECISION MAKING PERFORMANCE

.

Definition of INDIVIDUAL VARIATION IN ACTION IDENTIFICATION

H2

:

H3

:

METHODOLOGY

PARTICIPANTS AND DESIGN

EXPERIMENT PROCEDURE

MANIPULATION & MEASUREMENT OF INDEPENDENT VARIABLES

MEASUREMENT OF DEPENDENT VARIABLES

RESULTS

RANDOMIZATION AND MANIPULATION CHECK

IOWA GAMBLING TASK RESULTS

APARTMENT TASK

DISCUSSION

LIMITATIONS

CONCLUSION AND IMPLICATIONS

REFERENCES

APPENDICES

Appendix A

Appendix B

Appendix C

Appendix D

Appendix E

Appendix F

Appendix G

_{Supervisor: Prof. Dr. Bob Fennis}