Time Scarcity Recall and Its Effect on Decision Making Quality

0

Time Scarcity Recall and Its Effect on

Decision Making Quality

Master Thesis

by

Asunción Castresana López

University of Groningen

Faculty of Economics and Business

Msc Marketing Management

Final Thesis (6998 words)

June 16

th

_{, 2016}

Weeshuisgang 4-17 9712 EV Groningen The Netherlands +31 633 770 391 Student number 2399660 m.a.castresana.lopez@student.rug.nl asuncastresana@gmail.comasun

1

Table of Content

Abstract ... 2

Introduction ... 3

Research framework ... 4

Time scarcity and time abundance ... 4

Priming ... 6

Agency Level: Behavioral Identification Form ... 7

Decision Making ... 8

Iowa Gambling Task ... 9

Theory of Unconscious Thought ... 9

Research Design and Methodology ... 10

Participants and Design ... 10

Procedure ... 11

Independent Variable ... 14

Primed Time: Scarcity vs. Abundance ... 14

Behavioral Identification Form (BIF) ... 15

Dependent Variables ... 16

Iowa Gambling Task ... 16

Dijksterhuis Apartment task... 17

General Variables ... 18

Results ... 18

Iowa Gambling task ... 19

Dijksterhuis Apartment task ... 23

Conclusions and Recommendations ... 26

Conclusions Iowa Gambling task ... 26

Conclusions Dijksterhuis apartment task ... 27

General discussion ... 27

Limitations and future research ... 28

Managerial implications ... 29

References ... 30

Appendix A – Analysis output ... 34

Appendix B - Survey ... 36

2

Abstract

Time pressure has been proven to affect decision making of people, which is the main point of interest in this research. The main question it seeks to answer is as following: does simply recalling prior time pressure affect decision making quality? To test this, two decision making tasks are used in this research in order to test for the effect of primed time pressure in subjects. The used tasks, involving real life decisions, have been collected from previous researches. The results indicate that decision making quality is indeed affected by the recall of prior moments in life in which time pressure was felt by the subjects.

At the same time, personal susceptibility to external factors (such as time pressure) might interact with the effect of priming and therefore it was expected to find an interaction effect between susceptibility and priming. While priming and personal susceptibility were proven to be significant and partially significant respectively when it comes to decision making, the interaction for both measurements did not meet the expectations and was found insignificant.

3

Introduction

For the last two years, I have combined my master with an internship at a technology company in Amsterdam. I had long traveling many days each week, leaving me little time for self-enjoyment. During this time, the mere thought of exams in the upcoming semesters, regardless of the fact that it just started, caused lots of pressure. This affected my behavior and the quality of my decisions, just by thinking of the little time I would have then for other things rather than studying. I was barely able to visit my family and even when I did, I

couldn’t relax due to the load of work and exams which were awaiting me, sometimes feeling even more pressured during the time prior to the exams, than during the actual exam period.

Now my studying period has nearly reached its conclusion, and I started wondering whether thinking about stressful times at the work place would affect my behavior and decision, as much as exams have done in the past with me, and also whether this effect only applies to me or if it can be generalized to a certain extent. An initial search in existing literature shows that multiple authors have previously shown that actual time pressure can lead to poor decision making, but does priming time pressure experience have the same effect?

In this paper I expect to find an ordinal interaction between decision making and personal susceptibility, when time pressure as an experimental condition is manipulated via priming using prior personal experiences. Subjects get exposed to a prime of an autobiographical experience in which they experienced time pressure or time abundance in order to see in which situation priming of time leads to better decision making (Figure 1).

4 The above graph is a visual representation of the expected interaction of agency levels and decision making with time scarcity as manipulation. The three hypotheses which this research intends to answer are based on this, and formulated as following:

H1: Priming subjects with time scarcity will lead them to make less successful decisions.

H2: Low level agents (low BIF) will perform poorer on decision making tasks, than high

level agents (high BIF).

H3: The effect of primed time scarcity on decision making will be stronger for low level

agents than for high level agents

Explanation and development in more detail will follow in further chapters. The purpose of this paper is to observe if priming situations in which respondents experienced time scarcity or abundance, have an effect on the quality of decision making. Literature about the effects of time pressure and decision making already exists, but no existing research has focused on the effects of the recalling of experiencing time pressure on our daily decisions.

The relevance of this paper for the economic field is such that it could help to understand some effects of marketing communication of companies on the behavior and decision processes of the message receiver, with time pressure inducing messages like op=op1, and

also to know how personal differences can lead to different outcomes in this respect. This paper could help to follow up on differences between cultures, and how some people might be more susceptible to the effect of time pressures. In the social field this research can help with figuring out if people are influenced by perceived time pressure in their daily life, that their decisions are less successful and further develop on this to use it as a marketing strategy.

Research framework

Time scarcity and time abundance

Time scarcity effects have often been discussed in prior researches, even though some of these results contradict each other (Compton, 2005). Many researches have established that if people feel like the needed resources to fulfill their needs are low or might become

5 unavailable, a scarcity mind set arouses that changes the way their decisions are made (Mani et al., 2013; Mullainathan & Shafir, 2013; Shah et al., 2012).

Another theory about how time scarcity can have an impact on decision making, is the reactance theory (Brehm, 1966). In this theory it is established that when a subject’s freedom gets threatened, a resistance state is adopted, which triggers behaviors which can solve the threat and recover freedom. When subjects get exposed to time scarcity, they feel pressured and are expected to rush in their decisions in order to finish with the activity which induced the scarcity, to regain their feeling of freedom.

According to Wright (1974), time pressure affects the way humans make decisions, by giving more importance to the negative outcomes. According to Janis (1983) the pressure from the lack of time can lead the respondents to make premature decisions without evaluating all the options, just to reach closure. Janis and Mann (1977) found out that time constraints have an effect on decision making by reducing its efficiency. Svenson (1992) in his Differentiation and Consolidation Theory, establishes that time constraints deplete the capabilities of the human mind, preventing the differentiation between the two best options and that conscious, complex decisions in which alternatives need to be compared are

significantly influenced by time pressure, while automatic unconscious decision processes are not.

Svenson and Benson (1993) state that “both objective deadlines and subjective appraisal of task demands and the ability to meet these demands should affect cognitive processes under time pressure”. In their paper, they induce an objective time pressure by indicating to respondents to perform a task in “more” or “less” time than usual, and its results differ from actual time pressure generation – by decreasing the time to perform the task – from no changes compared with actual time decrease, to inducement of time pressure, or coping behavior for the “lack of resources”. The research proved that a change between “sufficient” and “too short” time stated in the instructions, does not make a difference in performance compared with real time pressure, choosing the same way of information processing for both statements. Unlike Svenson and Benson (1993) the author of this paper believes that priming a prior personal experience of time pressure, recreates the effects of genuine time pressure due to the personal and vividness of the experience.

6 Because of it, the researcher decided to present subjects with a priming stimulus of “time abundance” rather than “sufficient time” as in the original research, in order to avoid possible multicollinearity.

Petty and Cacioppo (1986) developed the elaboration likelihood model (ELM), which explains how attitudes are created or changed. The ELM explains how different routes to process stimuli information can lead to different outcomes. The central route they describe involves high levels of cognitive effort to process the arguments the subject is being exposed to, while the peripheral route relies on other cues than rational arguments, such as heuristics. When exposed to time scarcity, subjects are expected to process information by means of the peripheral route, meaning that the decision will be based on general evaluations rather than the rational. Unlike this, abundance exposed subjects are expected to make a more rational decision based on the information presented, and make use of the central route of information processing.

Priming

In order for the subjects participating in this research to experience time pressure or time abundance, different ways have been considered of recreating pressure, such as imagining a given situation, recreating conditions which would lead to such a situation of time pressure or priming a prior experience in which the subjects found themselves in similar situations.

Priming is defined as “the incidental activation of knowledge structures, such as trait concepts and stereotypes” (Bargh et al., 1996) which can have effect on behavior by

introducing subtle cues (primes) in the environment (Garcia et al., 2002). Bargh et al. (1996) state that priming studies are concerned with the effects that a current situational context and the environment cause individuals to think, feel, and behave differently than otherwise, hence how do subjects think, feel and behave differently when they are exposed to one time prime or another (scarcity or abundance). Based on the human memory theories, priming is defined as the activation of stored nodes in the memory network, which are related to other concepts via semantic paths related to external stimuli. When this node is activated, it can work as a filter to interpret and judge information (Collins & Loftus, 1975; Higgins & King, 1981; Taylor & Fiske, 1978; Wyer & Srull, 1981; Pan & Kosicki, 1997). Iyengar and Kinder (1987)

7 According to Suengas and Johnson (1988), memory based experiences are clearer and stronger than imagined ones, meaning that for a more intense and direct effect of the time pressure, it is better to recall an actual experience in the memory of the subjects than asking them to create a described situation with their imagination. These arguments were sufficient to decide for using priming prior experiences with respondents in order to be most accurate.

According to Nisbett and Ross (1980) priming prior experiences make the memories of those more vivid in the respondent’s mind, which makes them feel more “proximate in a sensory temporal or spatial way”, therefore experiencing the feelings of the primed experience in the current situation.

With these arguments, we can assume that when time scarcity is primed (vs time

abundance) at the beginning of the experiment, the pressure stemmed from it will negatively affect the decisions made afterwards, and therefore the main effect researched in this research is:

H1: Priming subjects with time scarcity will lead them to make less successful

decisions.

Agency Level: Behavioral Identification Form

Personal differences have a saying in how humans process information and end up choosing one option over another, and how susceptible we are to external factors. Vallacher and Wegner (1989) addressed this difference in susceptibility by creating a model of levels of personal agency, by identifying the differences, that subjects have to describe an action via the Behavioral Identification Form (BIF). The Action Identification Theory establishes that two different identities exist in a cognitive hierarchy; low-level identities which specify how one acts and high level identities that specify the reason to act or with what effect one acts (Vallacher & Wegner,1989). The scale allows researchers to assign individual respondents to one level or another of agency (high or low), depending on the goal orientation of their action statements. High level agents are expected to perform significantly better than low level agents when it comes to decision making.

8 when priming time scarcity, those individuals with high personal agency levels (high BIF score) will not be affected by the thought of lack of time when making a decision, but those with low personal agency levels will be affected by this. In the case of this research, high level agents are expected to neglect the influence of the manipulation – time pressure – in order to more successfully reach the goal of the performing tasks.

H2: Low level agents (low BIF) will perform poorer on decision making tasks, than

high level agents (high BIF).

H3: The effect of primed time scarcity on decision making will be stronger for low

level agents than for high level agents.

Figure 2: Conceptual Model

Decision Making

Decision making is a concept which was highly studied during the beginning of the second half of the twentieth century, and many definitions have derived from these researches.

According to several authors (Raiffa, 1968; Beyth-Marom et al., 1991), decision making is the process of making choices among competing course of actions which involves “cognitive processes, such as information search, evaluation, judgment and problem solving, as well as responses to a set of motivational forces that determine the manner in which decisions are made” (Mann et al., 1991).

9

Iowa Gambling Task

The Iowa Gambling task (IGT) was developed by members of the department of Neurology of the University of Iowa for their paper Insensitivity to future consequences

following damage to human prefrontal cortex (1994) and since then has been widely used to

measure affective decision making (Cauffman et al., 2010).

IGT subjects are presented with four decks of card of which they have to select one at a time, from which after the selection, they will get a reward or a penalty. Two of the decks (A and B) have a higher reward than the other two (C and D), but at the same time the penalty’s from those are also higher. Alike in real life, subjects cannot make calculations of the

outcomes, but they can make estimations of what would in the long run be the most

profitable, by assessing the risk of their decisions. The research concludes that subjects of the study must rely on their own ability to estimate which decks are risky and which are

profitable in the long run.

IGT researchers found in their study that age and gender have an influence on

performance, with older males having a significantly better performance, which is in line with previous researches (Cauffman et al., 2010; Crone et al., 2005; Crone and van der Molen, 2004; Crone, Vendel and van der Molen, 2003; Hooper et al., 2004).

De Ridder and colleges (2014) argued that gambling while on an empty stomach after fasting overnight were better at complex decision making. This research used IGT to assess quality of decision making when a resource (food) is scarce. Unlike the positive results of scarcity on decision making found in this research, the expected effect of time scarcity on decision making in the following one is negative.

Theory of Unconscious Thought

Dijksterhuis established across several researches among colleagues (Dijksterhuis, 2004; Dijksterhuis et al., 2006; Dijksterhuis & Meurs, 2006) that unconscious thoughts lead to improved decision making in complex situations while conscious thought is better in simpler decisions.

In his paper Think different: The Merits of Unconscious Thought in Preference

development and Decision Making (2004) Dijksterhuis defines Conscious Thought as “the

10 to a task”, while Unconscious Thought refers to “cognitive and/or affective task-relevant processes that take place outside conscious awareness.” The outcome thought of a decision is conscious but the process in between indecision and preference over an option is resulting from unconscious thought. Conscious thought is good at focusing on something but it neglects the less obvious or accessible information while Unconscious thought is more associative and reaches this less obvious and accessible information (Dijksterhuis & Meurs,2006).

The research established the expected result that conscious subjects “would not be able to take into account all the information and therefore focus only on a few attributes” unlike the unconscious subjects who we expected not to suffer any problem, being “easier for

unconscious thought to form a more global judgement based on all information”.

We can relate unconscious thought with time abundance, since the lack of pressure leaves the subject to focus on the non-obvious information and make a more thorough evaluation, while the time scarce exposed subjects will likely focus on some more general, obvious details and not make a well-balanced evaluation.

Because of these arguments, the expected results for the Dijksterhuis task are that those subjects primed with time abundance will make a better overall evaluation of the options, by being able to assess one option as the most desirable one, unlike those primed with time scarcity who would focus on some part of the information and won’t make an overall good decision based on the information.

Research Design and Methodology

Participants and Design

In order to find participants, students and employees were approached in the public places at the University of Groningen, who voluntarily joined the research (43 male, 53 female;

MAge= 27.81; SD= 9.503). One hundred respondents participated in the experiment; 50 in

11 Figure 3: Research Design

Procedure

Respondents were approached to participate in a MSc. experiment, in which two small tasks have to be conducted. These tasks take a couple of minutes to complete, with the help of a laptop and a tablet. On the laptop the participants were provided with a questionnaire

through Qualtrics. After this they were presented with a tablet on which they completed part of the experiment by means of the app “Iowa Gambling task” for the homonymous task. Respondents were assigned in advance with a participant number, which they entered in both devices prior to the start of the experiment in order to connect IGT scores with the rest of the experiment.

As introduction (Image 1) to the experiment, respondents were presented with the laptop, in which a brief description of what they are participating in was shown. The full experiment transcript can be found in Appendix B.

12 After this introduction, respondents are primed either with time scarcity or time abundance (50 of each), the independent variable, by asking them to recall prior experiences of time scarcity or abundance and requesting them to describe these as detailed as possible, as well as to indicate how this experience made them feel (Appendix B:pages 3A and 3B).

Once these prior time stimuli have been primed, respondents are asked to participate in a gambling task. They are introduced to the task both by the experiment and by the application (Image 2), clarifying that the participation’s goal is to maximize a fictitious amount of money by gambling with card decks (Appendix B: Page 4 and Ipad2). The respondent number was entered in the tablet, the same as in the questionnaire, and the tablet was handed back to the respondents who are once again briefed, in more detail, about the task they are about to perform.

Image 2 Introduction to IGT

13 Image 3

Introduction to apartment task

After both tasks have been completed, the agency levels measures was introduces (BIF), for which respondents were requested to point out which of the two presented options better describes a behavior according to them (Appendix B:page 19).

To conclude the experiment, respondents are requested to indicate gender, age, and mood, as demographic data (Image 4)

14

Independent Variable

Primed Time: Scarcity vs. Abundance

Subjects were asked to recall a few previous experienced (Image 5) situations in which they had plenty of time (control group) or a lack of time (experimental manipulation). They are requested to mention, describe and express what kind of feelings these situations

provoked, in order to make these memories more vivid. Image 5 and 6 Priming stimuli

15 With these arguments in mind, we can assume that when time scarcity is primed at the beginning of the interview, the pressure stemmed from it will negatively affect the decisions made afterwards, compared with time abundance participants, who will perform better in the subsequent tasks.

Behavioral Identification Form (BIF)

Personal differences determine the method of information processing of each individual. With their Action Identification Theory Vallacher and Wegner (1989) addressed this issue in consumer behavior research by creating a model of levels of personal agency. By means of the Behavioral Identification Form (BIF) differences between subjects can be identified by the way these describe an action. In a cognitive hierarchy two types of identities exist; low-level identities which specify how one acts and high level identities that specify the reason to act or with what effect one acts (Vallacher & Wegner,1989). Respondents have to choose between two ways of describing an action such as:

Attending class:

- Sitting in a chair

- Looking at the blackboard

There are in total 25 sets of actions, which will give information about the respondents’ agency level. In order to assign each respondent to an agency level –1 or 0– a median split of all respondents was calculated, which resulted into median value of 18. Furthermore, a test for reliability of the scale was conducted, concluding a highly reliable scale with an α of .805. Each of the respondents were then assigned according to this median; above this score to 1:high level and below it to 0:low level. In this way there is an inter-sample determination of what is considered a high (scores above the median) and a low (scores below) score regarding BIF.

16 the difference in decision making between the two different agents, with a poorer decision making for low level agents compared with high level agents.

Dependent Variables

In order to measure the quality of decision making and test for the main effect hypothesis, two different tasks, used as experimental research and developed in prior literature, have been selected. Since the effects and measures differ in scales, the two sub-hypotheses were

developed in order to be capable of measuring and assessing the effect of each of the tasks. These tasks will be introduced in the following pages.

Iowa Gambling Task

The Iowa Gambling Task (IGT) was developed by Bechara and colleagues in 1994 and has been widely used to measure affective decision making since then.

The task consists of four decks of cards, and 100 try outs in which respondents have to maximize their profits. Each of the decks differs in reward and penalty amounts, and

respondents are unaware of how many try outs they have to maximize their rewards. Two of the decks have higher rewards and higher penalties than the other two decks, and therefore are considered the risky options. Calculations as in real gambling are not possible, so only

estimations of what is most profitable in the long run can be made. Bechara et al. (1994) state that subjects of the study must rely on their own ability to estimate which decks are risky and which are profitable in the long run.

Based on this information and the effects of time scarcity on decision making, the

outcomes of the IGT are expected to be better for those subjects who have been primed with time (abundance) instead of scarcity, because the latter ones would be pressured to make a decision faster due to the recall of time pressure, and therefore, won’t be able to assess correctly risk and profitability.

H1A: Priming subjects with time scarcity will lead them to take higher risks, and get an overall worse performance outcome in terms of earnings.

With a starting loan of $2000, subjects have to maximize their earnings, by identifying the least risky options which in the long run lead to loses rather than earnings. Decks A and B are considered the risky options while C and D the safe options. The safe decks had least frequent and lower penalties ($50-250), while the risky decks had a more frequent penalty and

17 respondents primed with time abundance, who would recognize the risky options and switch towards the safe gaining options, generating the most earnings. The expected outcomes are higher revenues and a choice of safe decks more often for the abundance subjects, over the lower revenues (or bigger losses) and overall higher number of risky decks choices.

Therefore, the measures used as dependent variable were Net Gains, Number of risky

decisions, and Number of safe decisions, which were given by the app (Image 7). Since

respondents had to choose hundred times a deck to gamble on, the percentages given by the app on the made choices were used as the absolute values for this measure –one percent equals one observation.

Image 7 IGT app outcomes

Dijksterhuis Apartment task

Based on the Theory of Unconscious Thought developed by Dijksterhuis and colleges, and the definition given in his paper 'Think different: The Merits of Unconscious Thought in Preference development and Decision Making'(2004), the differences between Conscious and Unconscious Thought are explained. Conscious thought regards the cognitive and affective processes of which subjects are aware during a task performance, while the latter concerns the unaware processes taking place during the task performance.

18 option(B) more desirable than the other three, with 8 positive and 4 negatives attributes

(Appendix B:Pages 6-17). Of the other three remaining options, one is considered “the worst” (D) with the most negative attributes(8) and only 4 positives, and two fillers (A and C) or neutral options are included with 6 positives and 6 negatives attributes each. Respondents are exposed one by one to each category attribute.

The measure for this task, is the perceived difference between B and D by the respondents. It is expected that respondents primed with time abundance will perceive B as the best and D as the worse, by grading B the highest and D the lowest score. Respondents primed with time scarcity won’t be able to assess the difference and therefore the scores will differ less between the best and the worse option, and therefore the sub-hypothesis for H1 was introduced:

H1B: Priming subjects with time scarcity will lead them to not be able to correctly assess the difference between best and worst apartment option.

Respondents’ perception regarding which option is the optimal one are expected to be influenced by the primed stimuli prior to the task. The expected result hereby is that those subjects who were primed with time scarcity are less able to identify the best from the worst choice due to the cognitive time pressure, unlike subjects primed with time abundance who will be better able to recognize B as the best apartment and D as the worst apartment. The ability to identify the best from the worst apartment is measured by the difference in score subjects give to B and D, whereby a larger positive difference means that subjects were successfully able to identify the best and the worst apartment.

General Variables

In addition to the experiment, variables of gender and age are asked at the end of, not interfering with the manipulation effects.

Furthermore, a general 1-7 mood scale was added as a control variable (Mmood:3.00;

SD:1.12), to certify in term of correlation, that the effects on the dependent variable are

indeed due to the manipulation administered at the beginning of the experiment, and not due to respondents’ mood.

Results

19 there is a significant difference in the mean values of the dependent variables across the manipulation of the independent variable.

In the experiment the independent variable was the priming of two specific time

circumstances, consisting of two levels: time abundance recall and time scarcity recall. The dependent variable of decision making was divided in two performing tasks: Iowa gambling task, and Dijksterhuis’ apartment task, and therefore two ANOVA’s were conducted to test the effect of the priming on both dependent variables.

After the data had been gathered, the data was explored to make sure that all results were valid and that people genuinely paid attention to the tasks they were performing. Therefore, a boxplot has been calculated in order to check for outliers. It is important to eliminate outliers, as these data points are extreme values, compared to the other values in the sample, and can significantly influence the analyses outcome. These extreme values can be due to errors, or non-serious participation by subjects. The boxplots in graphs 1 and 2 in Appendix A show how most people scored on certain variables, and which scores are outside the ‘normal’ range (those outside the lines). Based on this test, multiple outliers were identified. 4 respondents were outliers based on their Apartment task performance, since those were the only ones who rated D much higher than B in the high-BIF group, and one was considered an outlier due to the extreme IGT performance score, which was way out of the boxplot range.

After inspecting the identified outlier’s data sets, the conclusion was that the respondents who were considered outliers on the apartment task scoring a negative difference between B and D (out of a normal range between 0-6), did not genuinely participate in the tasks. These subjects filled in data with patterns such as 1,2,3,4 and were excluded from the data set. The respondent gaining almost 2000 points on the IGT task (out of the normal range between 2000 & -1000) must have been fortunate to avoid almost all penalties and the extreme score

influenced the data set too much. Further analyses included the remaining 95 respondents. Out of these respondents, after calculating the mean of the BIF scores, each was assigned with an agency level; 0 for those with a mean below 0.5 and 1 for those with a higher value, with a total of 46 low level agents and 49 high level agents.

Iowa Gambling task

20 and so is the range of scores, with much lower minimum scores in the scarcity group.

Regarding the amount of safe versus risky decks – in a scale from 1 to 100 – the abundance group had a higher amount of safe choices with a lower standard deviation. Also, it is relevant to notice that in the scarcity group, a higher maximum of risky choices was reached, with 93 out of 100 risky choices.

Table 1

Descriptives of the IGT Performance

When testing the main effect of time scarcity on decision making, an initial ANOVA to test the significance of the discrepancies between groups, showed (Appendix A: Table 8) that all 3 measures of performance of the IGT are significantly influenced by whether the

participant was primed with time scarcity or time abundance (Net gains F (1,93) = 19.64,

p = .000; Safe choices F (1,93) = 6.47, p = .013; Risky choices F (1,93) = 6.57, p = .012),

because all measures have a p-value of <.05. It can be seen that in all three measures, the scarcity group performed worse (Net gains M = -343, SD= 806.65; Safe choices M = 54.02,

SD = 15.35; Risky choices M = 46.18, SD = 15.83) than the control group primed with time

abundance (Net gains M = 357, SD= 725.82; Safe choices M = 61.64, SD = 13.68; Risky choices M = 38.36, SD = 13.68)

21 test showed that the found effect comes exclusively from the priming, and that BIF and the interaction effect of priming*BIF are not significantly influencing the performance. For the amount of safe and risky decisions the analysis shows the same conclusion: only priming affects negatively (Net Gains: p= .000; Safe choices: p= .014; Risky choices: p= .013) performance significantly, and increases the amount of risky choices made. These effects are shown below in tables 2-4 and graphs 1-3.

Table 2

Interaction effect on IGT net gains

Graph 1

Priming effect on IGT net gains

F P (<0.05)

BIF 1.70 .196

PRIMING 19.46 .000

22 Table 3

Interaction effect on safe choices

Graph 2

Priming effect on safe choices

Table 4

Interaction effect on risky choices

23 Graph 3

Priming effect on risky decisions

Dijksterhuis Apartment task

The measure used for this task to assess whether participants made good quality decisions, was their ability to successfully identify apartment B as a more desirable option than D. The perceived differences between options B and D in the apartment task, on a scale of 1 of 7, between control group and abundance group differs in the fact that the minimum of the control group is 0 –no difference was perceived– while in the manipulation group the minimum score is -3, meaning that some of the respondents identified D as a more desirable option than B (Table 5).

Table 5

Descriptives of Aparment task performance

The ANOVA analysis for the task showed that priming does have an effect (p=.064) in the perceived difference between a good and a bad choice only at a higher significant level of

p>.1, and therefore this level would be used as a threshold for all consequent apartment task

24 the difference between B and D (M = 2.98, SD= 1.56) than the respondents primed with time scarcity (M = 2.2, SD = 2.36).

After proving that a direct effect exists between the independent and dependent variable, a test was conducted (Appendix A: Table 11) in order to assess if the moderator had a direct effect on the dependent variable, which concluded that a significant effect exists between the BIF scores and apartment performance (F(1,93)=4.23, p=.043).

Due to the existence of this effect, and the direct effect between priming time scarcity and the performance of the task – at a lower significant level – it is necessary to check if a

significant mediation effect exists, through priming and agency levels. This analysis shows that there is no significant relationship between priming and agency levels (F (1,93) = 1.46, p

=.231), thereby excluding the possibility that the main effect between priming and

performance of the task is mediated by agency levels (Appendix A: Table 12).

A full factorial ANOVA corroborated that both priming and agency levels have a direct effect on the apartment task scores (BIF: p= .053; Priming: p= .065), while there was no interaction (p= .333) between those two variables affecting the dependent scores as it can be observed in Table 6 and the different effect caused by priming in both groups in graphs 4 and 5.

Table 6

Interaction effect on apartment task

F P (<0.1)

BIF 3.86 .053

PRIMING 3.50 .065

25

Graph 4 Graph 5

Priming effect on apartment task BIF effect on apartment task

Furthermore, a test was conducted to check if the influence of the personal mood at the moment of the research, gender and age had an influence on the performance of the tasks, by means of a correlation analysis. This analysis shows that no significant correlation between age, gender or mood with the performance measures of both IGT and apartment task were found; as can be seen below in Table 7.

Table 7

26

Conclusions and Recommendations

The analysis conducted to test the effect that factors like gender age and mood have on the quality of decisions made proved that no significant effect can be deducted from them. Furthermore, the rejection that age and gender have an effect contradicts previous researches (Schubert et al., 1999; Finucane et al., 2005; Graham et al., 2002) which stated that older adults rely more on their experience while making decisions, and woman are found to be more risk adverse in their decisions. These contradicting results might be due to the mainly below 30 years old sample, something which would be recommended to check for in further research.

Conclusions Iowa Gambling task

Figure 4 IGT hypotheses

The conducted test proved that priming a lack of time had a negative effect on the

performance of the IGT, represented as A1 in the conceptual map. Three measures were used

in order to assess performance: net gains, amount of safe choices and amount of risky choices, and all of these measures supported H1: Priming subjects with time scarcity will lead them to make less successful decisions, with priming lack of time provoking lower scores and higher

amount of risky decisions.

When testing the influence of agency levels on the model, represented as C1, no effect on

the performance was found out, rejecting H2: Low level agents(low BIF) will perform poorer on decision making tasks, than high level agents(high BIF). When testing for the interaction

effect, represented as B1, the results were insignificant, again confirming that all the effect

27

time scarcity on decision making will be stronger with low level agents, compared with high level agents.

Another conclusion is that no mediation effect was present, with an insignificant direct effect of agency level on the performance.

Conclusions Dijksterhuis apartment task

Figure 5

Apartment task hypotheses

Priming lack of time has a negative effect on performance of the apartment task, represented as A2, as well as with the IGT performance and therefore supports H1, with

priming lack of time provoking poorer performance on the task. Unlike in the IGT test, the agency level has a significant effect in the model by directly influencing the performance of the task, represented as C2 and therefore supporting H2.

However, the interaction between priming lack of time and levels of agency has no significant influence on the performance, represented as B2, ruling out H3.

General discussion

28 effect of it in the subsequent decisions made. The mere thought of moments of no time

induced feelings of stress in most of the respondents.

Belayachi and Van der Linden(2013) established that low level agents have difficulties understanding goal oriented actions, and so it has been proven to be for the apartment task as the goal was to figure out what apartment was the best. These results are in line with their research, where participants performed better at a lexical task after exposing them to a goal oriented scenario (Belayachi & Van der Linden,2013). With the apartment task, respondents were exposed to the briefing explaining the goal of the task, to find the best apartment, and later exposed to lexical information of apartments. As theoretically predicted, the results showed that the higher the BIF, the better respondents performed when attempting to identify the difference between apartments B and D. The authors’ work also explains why in this research agency levels did not influence the performance of the gambling task, since it did not involve a lexical processing task, in which high level agents would have performed better.

Unlike was expected, no interaction effect between agency levels and priming time pressure was found, confirming that individually both variables are significantly influencing the decision making process humans undergo.

Limitations and future research

The respondents of this research were mainly young, highly educated people, since they were gathered at the university common areas and at the marketing department of an

international company, therefore it would be useful to reproduce this research with different demographics, to be able to better identify underlying possible differences between groups.

When conducting the apartment task, some participants notified the researchers of the fact that they focused predominantly on the option which they perceived as the best, and did not pay much attention to the rest of options after that and therefore gave bad scores to those which were not included in their initial consideration set.

One more limitation was the fact that the research was conducted in English and most of the persons who participated were non-native English speakers, and a smaller, but

29

Managerial implications

With the provided insights that priming lack of time affects the overall quality of decisions can be useful for marketing purposes. Following the results obtained from this research, marketing managers can use these findings to lead consumers towards a favorable decision. Marketing messages pursuing consumers to make rational choices, based on an overall evaluation, should allow them to think of moments with no time pressure in order to identify the superiority of the proposition. On the other hand, messages focusing on specific aspects of products could prime moments with time scarcity in order for consumers not to make an overall evaluation and consider only the prominent feature to evaluate the likability of a product.

The influence of agency levels–personal susceptibility has been supported only partly by the research, and further research is advised to find out how marketers can benefit by

30

References

Bargh, J. A., Chen, M., & Burrows, L. (1996). Automaticity of social behavior: Direct effects of trait construct and stereotype activation on action. Journal of personality and

social psychology, 71(2), 230-244.

Bechara, A., Damasio, A. R., Damasio, H., & Anderson, S. W. (1994). Insensitivity to future consequences following damage to human prefrontal cortex. Cognition, 50(1), 7-15.

Belayachi, S., & Van der Linden, M. (2013). Individual differences in cognitive representations of action influence the activation of goal concepts. Acta

psychologica, 142(2), 259-264.

Beyth-Marom, R., Fischhoff, B., Quadrel, M. J., & Furby, L. (1991). Teaching decision making to adolescents: A critical review. In Baron, J. & Brown, R.V. (Eds.),

Teaching decision making to adolescents (p. 19-59). New York: Routledge.

Brehm, J. W. (1966). A theory of psychological reactance. New York: Academic Press

Cauffman, E., Shulman, E. P., Steinberg, L., Claus, E., Banich, M. T., Graham, S., & Woolard, J. (2010). Age differences in affective decision making as indexed by

performance on the Iowa Gambling Task. Developmental psychology, 46(1), 193-207. Collins, A. M., & Loftus, E. F. (1975). A spreading-activation theory of semantic processing. Psychological review, 82(6), 407-428.

Compton, J. A. (2005). Rising to the Challenge or Cracking Under Pressure? Time Scarcity & Effects on Performance in Limited Preparation Events. Forensic, 90(2), 1-15.

Crone, E. A., & van der Molen, M. W. (2004). Developmental changes in real life decision making: Performance on a gambling task previously shown to depend on the ventromedial prefrontal cortex. Developmental Neuropsychology, 25(3), 251–279.

Crone, E. A., Bunge, S. A., Latenstein, H., & van der Molen, M. W. (2005).

31 Crone, E. A., Vendel, I., & van der Molen, M. W. (2003). Decision-making in

disinhibited adolescents & adults: Insensitivity to future consequences or driven by immediate reward? Personality and Individual Differences, 35(7), 1625–1641.

De Ridder, D., Kroese, F., Adriaanse, M., & Evers, C. (2014). Always gamble on an empty stomach: Hunger is associated with advantageous decision making. PloS

one, 9(10), e111081, 1-9.

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.

Dijksterhuis, A., & Meurs, T. (2006). Where creativity resides: The generative power of unconscious thought. Consciousness and cognition, 15(1), 135-146.

Dijksterhuis, A., Bos, M. W., Nordgren, L. F., & Van Baaren, R. B. (2006). On making the right choice: The deliberation-without-attention effect. Science, 311(5763), 1005-1007.

Finucane, M. L., Mertz, C. K., Slovic, P., & Schmidt, E. S. (2005). Task complexity and older adults' decision-making competence. Psychology and aging, 20(1), 71-84.

Garcia, S. M., Weaver, K., Moskowitz, G. B., & Darley, J. M. (2002). Crowded minds: the implicit bystander effect. Journal of personality and social psychology, 83(4), 843-853.

Graham, J. F., Stendardi, E. J., Jr., Myers, J. K., & Graham, M. J. (2002). Gender differences in investment strategies: An information processing perspective. International

Journal of Bank Marketing, 20(1), 17-26.

Higgins, E. T., & King, G. (1981). Accessibility of social constructs: Information processing consequences of individual and contextual variability. In Cantor, N. & Kihlstrom, J.F. (Eds.), Personality, cognition, and social interaction (p. 69-121). Hillsdale, NJ: Erlbaum.

32 making and ventromedial prefrontal cortex. Developmental Psychology, 40(6), 1148– 1158.

Iyengar, S., & Kinder, D. R. (1987). News that matters: Television and American

opinion. University of Chicago Press.

Janis, I. L. (1983). The role of social support in adherence to stressful decisions. American Psychologist, 38(2), 143-160.

Janis, I. L., & Mann, L. (1977). Decision making: A psychological analysis of conflict,

choice, and commitment. Free Press.

Mani, A., Mullainathan, S., Shafir, E., & Zhao, J. (2013). Poverty impedes cognitive function. Science, 341(6149), 976–980.

Mann, L., Harmoni, R., & Power, C. (1991). The GOFER course in decision making. In Baron, J. & Brown, R.V. (Eds.), Teaching decision making to adolescents (p. 61-78). New York: Routledge.

Mullainathan, S., & Shafir, E. (2013). Scarcity: Why having too little means so much. New York, NY: Times Books.

Nisbett, R. E., & Ross, L. (1980). Human inference: Strategies and shortcomings of

social judgment. Englewood Cliffs, NJ: Prentice-Hall.

Pan, Z., & Kosicki, G. M. (1997). Priming and Media Impact on the Evaluations of the President's Performance. Communication Research,24(1), 3-30.

Petty, R. E., & Cacioppo, J. T. (1986). Communication and Persuasion: Central and Peripheral Routes to Attitude Change. New York: Springer.

Raiffa, H. (1968). Decision analysis: Introductory lectures on choices under

uncertainty. New York: McGraw-Hill.

33 Shah, A. K., Mullainathan, S., & Shafir, E. (2012). Some consequences of having too little. Science, 338(6107), 682–685.

Shah, A. K., Shafir, E., & Mullainathan, S. (2015). Scarcity frames value.

Psychological science, 26(4), 402-412.

Suengas, A. G., & Johnson, M. K. (1988). Qualitative effects of rehearsal on memories for perceived and imagined complex events. Journal of Experimental

Psychology: General, 117(4), 377-389.

Svenson, O. (1992). Differentiation and consolidation theory of human decision making: A frame of reference for the study of pre-and post-decision processes. Acta

Psychologica, 80(1-3), 143-168.

Svenson, O., & Benson III, L. (1993). On experimental instructions and the

inducement of time pressure behavior. In Svenson, O. & Maule, J. (Eds.), Time pressure

and stress in human judgment and decision making (p. 157-165). New York: Springer US.

Taylor, S. E., & Fiske, S. T. (1978). Salience, attention, and attribution: Top of the head phenomena. Advances in experimental social psychology, 11, 249-288.

Vallacher, R. R., & Wegner, D. M. (1989). Levels of personal agency: Individual variation in action identification. Journal of Personality and Social Psychology, 57(4), 660-671.

Wright, P. (1974). The harassed decision maker: Time pressures, distractions, and the use of evidence. Journal of applied psychology, 59(5), 555-561.

34

Appendix A – Analysis output

Graph 6 Outliers IGT

Graph 7

35 Table 8

Effect of priming on IGT measures

Table 9

Moderating effect of BIF on IGT

F _P<0,05

Net Gains IGT _{1.86 .176}

Safe Decks .28 .596

Risky Decks .21 .648

Table 10

Effect of priming on Apartment task

Table 11

Effect of BIF on apartment task performance

Table 12

Effect of Prime on BIF

36

Appendix B - Survey

Page 1 – Respondent number

37

38

39

Page 4 – Introduction to IGT

40

41

Ipad (3) Gambling Task

42

Change to Laptop – Page 5 Introduction to apartment task

48

49

53

Page 20 – Demographics and mood

54

Appendix C – IGT App and BIF form

Images 8 and 9

55 Image 10