Can individual differences in decisiveness moderate the choice overload effect? They could under conditions that elicit a choice overload

Student: Zoi Portokalidou Student number: s2100630

First examiner: Dr. Ruigendijk H.A.H. Second examiner: Dr. Pliskin R.

Can individual differences in decisiveness moderate

the choice overload effect? They could under conditions

Abstract

The choice overload effect illustrates a decision-making phenomenon in which individuals tend to experience lower satisfaction with their chosen item when they choose from a large set of options than from a small one due to anticipated regret and cognitive overload (Iyengar & Lepper, 2000; Schwartz, 2004). The present study investigated the joint effects of choice overload and individual differences in decisiveness on satisfaction with selected partners in an online-dating situation. 84 single Dutch males and females were randomly assigned to either an extensive choice (24 profiles) or a limited choice (6 profiles) condition and chose a dating partner. Subsequently, they reported their satisfaction with the specific partner. A regression analysis demonstrated a significant but reversed effect of quantity of choice condition. Surprisingly, participants in the extensive choice condition reported higher satisfaction than those in the limited choice condition. Contrary to our hypotheses, decisiveness and the

interaction between decisiveness and quantity of choice condition did not significantly predict satisfaction with the selected partners. However, it is questionable whether participants of the present study indeed experienced a choice overload effect or not. Implications of when online-dating situations can elicit a choice overload effect are discussed.

Keywords: online dating, decisiveness, choice overload effect, action-orientation,

Table of Contents

Introduction ... 3

The Choice Overload Effect ... 3

Decisiveness: Action versus State-Orientation ... 6

The present study ... 9

Method ... 11 Participants ... 11 Design... 11 Procedure ... 11 Materials ... 13 Results ... 14 Discussion ... 16 References ... 25 Appendices ... 29

Appendix A: Cover Story ... 29

Appendix B: Debriefing Form ... 31

During the past decade online-dating is considered an improvement and to a certain extent even a displacement of traditional dating. The increasing popularity of online dating is due to a recent decrease of the social stigma associated with it and to its time-saving nature (Ellison, Heino, & Gibbs, 2006). Although many studies propose that relationships formed online tend to become stable over time and end up in satisfactory marriages (McKenna, Green, & Gleason, 2002; Weiten, Dunn, & Hammer, 2014), there are also obstacles associated with online dating. One of those is the abundance of available profiles which can affect evaluations regarding the selected partner (Finkel, Eastwick, Karney, Reis, & Sprecher, 2012).

The purpose of the present study is to investigate the satisfaction with partner selection in online-dating, while taking into account the effects of the amount of dating profiles and decision-related individual differences. We focus on individual differences in decisiveness because online-dating is unavoidably linked to decision-making processes. Additionally, the focus on the amount of available options affecting satisfaction is inspired by previous evidence from the consuming field of psychology and has been referred as “the choice overload effect” (Schwartz, 2004). Several implications of this effect and why individual differences might be important within this context will be discussed.

The Choice Overload Effect

The choice overload effect contradicts the commonsense belief that a plethora of options can be beneficial because an increase in control and variety lead to the best possible choice. Paradoxically, individuals tend to experience lower satisfaction with their choices when they choose from a large pool than from a small one (Schwartz, 2004). Iyengar and Lepper (2000) initially investigated how individuals deal with amount of options through comparisons of participants that had to choose jams, chocolates or to undertake assignments. Although large assortments were associated with higher enjoyment and more initial attraction, participants who were assigned to large pool conditions (i.e. 24 jams) reported more dissatisfaction and less motivation to purchase such items compared to those who were assigned to small pool

conditions (i.e. 6 jams). Later studies demonstrated comparable results and extended the choice overload effect to other items and conditions, including high-stakes contexts. Such contexts are, for instance, interpersonal relationships and mate selection (Lenton & Francesconi, 2010; D’Angelo & Toma, 2017).

Different models and theories of cognition and behavior have attempted to identify the reasons behind the choice overload effect. The most common explanation provided for this effect is anticipated regret that is stimulated by options’ increase. The regret that one might experience in the future for the choice he or she is making decreases excitement for the

selected item (Iyengar & Lepper, 2000). However, the number of options does not affect regret only directly. The evaluation of rejected options’ values and characteristics creates

counterfactual thinking and regret about the chosen item. Namely, individuals tend to think through the rejected alternatives’ characteristics after making their choice. This process, in turn, creates doubts and decreases excitement about the final choice, as they tend to wonder “what could have been” if they had chosen differently (Sagi & Friedland, 2007).

Along with that, evaluating all options to make a decision is a complex process that creates a cognitive load (Haynes, 2009). This complexity can be enhanced when options in set sizes have many variable attributes. Processing many attributes of many available options creates attentional and memory loads that in turn affect the quality of decision and the choice strategy (Reutskaja & Hogarth, 2009). Specifically, choices in large sets are usually induced by quickly and easily assessed cues and by heuristic-like strategies since time for processing and cognitive resources are limited (Lenton & Francesconi, 2010).

Choice justification also seems to explain to some extent this paradoxical effect. Situations in which justification for the choice is required have been shown to prompt a choice overload effect (Scheibehenne, Greifeneder, & Todd, 2009). Because individuals might need to

justify to others why they discarded other good alternatives, they tend to experience decreased excitement and satisfaction with their final choices (D’Angelo & Toma, 2017).

Since dating involves the process of selecting a partner among other available ones, the choice overload effect can be applied to mate selection in an online-dating context. For

example, online daters report that they prefer to have the option to browse more profiles when searching for partners online, although they are aware of the threats associated with such an increase of available options (e.g. dissatisfaction) (Lenton, Fasolo and Todd, 2008). However, they are not willing to engage in profile searching for a long time when they are assigned in large pool conditions (20 profiles) compared to when they are assigned to small ones (4 profiles). Furthermore, as set sizes increases, online daters tend to engage in extensive profile searching and in turn make poorer choices based on their reported characteristics for ideal partners (Wu and Chiou, 2009). Increase in partners’ features generates more confusion among daters that leads to heuristic-like processing strategies and then to poorer choices. Specifically, when time and cognitive resources are limited, individuals do not engage in deep processing but they are rather judging by easily-assessed cues (e.g. weight). As a result, these selection processes might not reflect their usual choice strategies. This can decrease their satisfaction with their choice because of the inconsistency with their usual attitudes of choosing (Lenton & Francesconi, 2010).

The most recent evidence on the choice overload effect in online-dating is a study by D’Angelo and Toma (2017) that measured the explicit satisfaction with participants’ partners. They created their own online-dating website for university students and had participants select dating partners with the excuse of testing the overall website’s structure that would be

launched soon. Participants were assigned to either an extensive choice (i.e. 24 profiles) or a limited choice condition (i.e. 6 profiles) and had to select only one dating partner. Indeed, individuals that had chosen from a large set reported lower satisfaction with their partner

compared to those that had chosen from a small one. They measured participants’ satisfaction with partners both immediately and after one week of selection. The specific timeframe was selected because they suggest that online-dating choice is a psychologically challenging situation that may cause dissonance. Dissonance in turn can trigger regret. Regret, as well as contemplation, needs some time to be processed and unfolded. Earlier comparable studies had not found such differences in satisfaction. This divergence might be because past studies measured the satisfaction right after participants chose partners. This study highlighted the explicit relationship between choice satisfaction and high-stakes contexts, opening the way for the extension of the choice overload effect in more complex environments.

Decisiveness: Action versus State-Orientation

Several processes underlying the choice overload effect differ between action and state-oriented individuals. Therefore, individual differences in action versus state-orientation were expected to moderate the choice overload effect. Such individual differences are defined by Kuhl’s (1994) action control scale that distinguishes action from state-oriented individuals (Kuhl, 1994). The differentiation between action versus state-orientation implies that

individuals’ self-regulation of emotional and motivational states can be processed within one of two volitional modes. Specifically, action orientation is associated with a metastatic mode that stimulates flexibility of self-regulatory process and state orientation is associated with a catastatic mode that stimulates fixation toward negative states (Beckmann & Kuhl, 1984; 1994). Action versus state orientation can moderate the effects of a demanding context on self-regulation processes. An increase in demands benefits action-oriented individuals by

stimulating self-regulatory resources. However, it harms state-oriented individuals due to their negative adjustment as a consequence of the elimination of the impact of high level goals and action-related intentions on behavior (Koole, Jostmann & Baumann, 2012). Although

individuals might switch between modes depending on situational factors, a chronic tendency towards one mode can indicate one’s personality trait. The demand-related action versus state

orientation dimension reflects decisiveness or hesitation under demanding conditions that is distinct from other personality traits (e.g. impulsivity).

Beckmann and Kuhl (1984) demonstrated that action-oriented individuals process available options in a way that promotes quick actions by ignoring irrelevant information. However, state-oriented individuals process all the available information equally - a tendency that creates confusion that impedes the execution of actions. To our knowledge, choices made in large set conditions are induced by heuristic-like strategies. At the same time, in dating websites there is high variability among available partners that can create cognitive confusion. Therefore, the ability to ignore irrelevant information and focus on action (e.g. decision) could benefit action-oriented individuals in an extensive choice condition.

Moreover, Jostmann and Koole (2006) investigated how working memory load differs depending on action versus state-orientation. They introduced to participants a relationship priming of either a demanding (high-demand condition) or an accepting (low-demand condition) visualized partner. Their results indicated that action-oriented participants scored higher in working memory capacity under demanding conditions compared to state-oriented. After such demanding priming, action-oriented individuals were also more likely compared to state-oriented individuals to show superiority intention memory effect when associating words in a recognition test with intention scripts they had read a priori. Those effects were not present in low-demanding conditions. Thus, action-oriented individuals efficiently utilize and update their working memory capacity under high-demanding situations, due to tendencies towards decisiveness and initiative that stimulate availability for subsequent tasks. The opposite applies to state-oriented individuals that show sustained working memory load due to inclination toward indecisiveness and hesitation. Therefore, in high-demanding situations that require cognitive effort such as a choice overload experience, state-oriented individuals might show higher working memory load that will maintain confusion. Contrariwise, action-oriented

individuals might utilize their capacity for subsequent tasks and focus on the goal of making a decision.

Importantly, action and state-oriented individuals differ in their experienced regret. After recalling a past action (going out) or inaction (staying home) situation, regret ratings were higher for inaction than action among action-oriented individuals. On the contrary, state-oriented individuals did not differ in their experience of regret between action or inaction situations. In both conditions their reported level of regret was high (McElroy & Dowd, 2007). This pattern is due to the inability of state-oriented individuals to efficiently regulate negative emotions. On the contrary, action-oriented individuals experience high regret only for

situations involving inaction because of the inconsistency with their orientation. In a dating context, selecting a partner is an action. State-oriented people tend to regret actions more that action-oriented ones do. As regret is an important underlying process of the choice overload effect, individual differences in action versus state-orientation could moderate it. In a choice overload situation, state-oriented individuals should experience more regret due to the

experienced difficulty of the situation (Iyengar & Lepper, 2002). Accordingly, action-oriented individuals should not experience high levels of regret if they select a partner, due to

consistency of the action with their orientation.

Moreover, action-oriented individuals compared to state-oriented ones have been shown to increase more the attractiveness ratings for their choices from a first to a second point of evaluation (Beckmann & Kuhl, 1984). Similarly, they tend to rate more positively their final choice compared to the discarded alternatives (Harmon-Jones & Harmon-Jones, 2002). Such findings imply that action-oriented individuals might be more inclined to display justification for their choices or to view them more positively. Therefore, in an extensive choice condition, action-oriented individuals might increase the perceived value of their choice and decrease it

for the numerous discarded alternatives. By that way, they could experience higher satisfaction and less regret with their selection compared to state-oriented individuals.

Therefore, given that a.) the process of choosing a partner is considered a “high-stakes context” because it affects well-being, individuals get emotionally involved and have to justify their choices to significant others (see D’Angelo & Toma, 2017); b.) variability is usually high in online-dating websites; and c.) the choice overload effect partially occurs due to limited cognitive resources, we considered online-dating process a demanding situation for individuals to deal with. This situation could be more demanding when individuals are confronted with a plethora of available partners. Therefore, demand-related action versus state-orientation could moderate the increase in demands in this context, due to differences in processing of options, post-decision behavior and anticipated regret.

The present study

The purpose of the present study was to investigate the effects of the choice overload and decisiveness based on individual differences in action versus state-orientation on

satisfaction with selected partners in online-dating. Satisfaction was measured both

immediately after the partner was selected and one week after that. The specific timeframe is because online-dating choice is a psychologically challenging experience that can cause dissonance and regret that need some time to be processed (D’Angelo & Toma, 2017).

Therefore, we expected that one week would be sufficient for action versus state-orientation to moderate the choice overload effect.

Specifically, given that in low-demanding conditions, action-orientation does not predict better self-regulation than state-orientation (Jostmann & Koole, 2006; Koole et al., 2012), we considered limited choice condition a low-demanding situation in which action and state-oriented individuals should experience the same choice satisfaction.

Hypothesis 1a: There will be no differences in satisfaction between action and state-oriented individuals in the limited choice condition.

Action-oriented individuals are better at regulating themselves under high-demanding conditions (Koole, Jostmann & Baumann, 2012); at ignoring irrelevant information

(Beckmann & Kuhl, 1984) and at updating working memory capacity (Jostmann & Koole, 2006). At the same time, they experience less regret after actions (McElroy & Dowd, 2007), and are more inclined to increase the attractiveness of their choices (Beckmann and Kuhl, 1984). Thus, we expected them to report higher satisfaction with their choice in a high-demanding situation.

Hypothesis 1b: Action-oriented individuals will experience higher satisfaction compared to state-oriented individuals in the extensive choice condition.

Accordingly, because online-dating context is an already demanding situation itself (D’Angelo & Toma, 2017) that can be more demanding by the increase of available options, we expected that action-oriented individuals would use their inclination toward initiative and action to deal with both conditions effectively.

Hypothesis 2a: There will be no difference in satisfaction scores for action-oriented individuals in the extensive choice condition compared to the limited choice condition.

On the contrary, state-oriented individuals show negative adjustment to the demands’ increase (Koole, Jostmann & Baumann, 2012), anticipated regret towards their actions

(McElroy & Dowd, 2007), high working memory load (Jostmann & Koole, 2006) and inability to ignore irrelevant information when making decisions (Beckmann & Kuhl, 1984). Thus, we proposed that state-oriented individuals would experience lower satisfaction when switching from a low-demanding to a high-demanding situation.

Hypothesis 2b: State-oriented individuals will experience lower satisfaction in the extensive choice condition compared to the limited choice condition.

Based on previous scientific literature about choice overload in online-dating contexts (D’Angelo & Toma, 2017), we hypothesized that in an online-dating situation, individuals

would indicate greater satisfaction when choosing from a small pool of options (6 profiles) than from a large one (24 profiles).

Hypothesis 3: Participants in the extensive choice condition will report lower satisfaction compared to participants in the limited choice condition.

In general, quantity of choice condition (limited versus extensive choice), decisiveness based on individual differences in action versus state-orientation, and their interaction, were expected to affect the satisfaction scores after one week of partner selection. At the same time, we expected that these variables will predict the satisfaction difference from immediate

selection to one week after that selection.

Hypothesis 4: Quantity of choice condition, decisiveness and their interaction will be significant predictors of satisfaction difference from Time 1 to Time 2.

Method Participants

Participants of the present study were initially 98. However, homosexual participants, participants younger than 20 and older than 28 years old, participants in a relationship, and participants who did not remember their selected partner after a week, were all excluded from the data. This exclusion left our sample with 84 single Dutch participants, from which 33 were males and 51 females, with an age range from 20 to 28 (M = 24.07, SD = 1.94).

Design

This study is a two-group between (extensive choice versus limited choice conditions) and within (Time 1 and Time 2) subjects design with individual differences in decisiveness as a moderator and satisfaction as the dependent variable.

Procedure

This study and its materials have been approved by the Psychology Ethics Committee of Leiden University. Participants were approached via social media and chatting applications. Initially, they were sent a message either on Facebook or on Whatsapp with the link of the

study. This link redirected them to the Qualtrics software. They were first presented with the informed consent which they had to sign by ticking a box to officially confirm their

participation to the study. The information sheet included a cover story (see Appendix A) which implied that this study is consisted of two different and independent researches.

Supposedly, the first study was consisted from two parts and was about testing a new website on dating (D’Angelo & Toma, 2017). The other one was about decisiveness as part of another independent research. After agreeing to participate and based on random allocation and on gender selection, participants screened either 6 (limited choice condition) or 24 (extensive choice condition) male or female profiles from which they were asked to select only one dating partner. Following their selection, they answered the satisfaction scale (Time 1). Subsequently, they were asked to type their e-mail and to write down the name of their chosen partner. They were also asked to note the date of the second part of the study in their calendar, as a reminder would be sent to them on that day. Then, they continued with a second irrelevant study about decisiveness. After answering the decisiveness scale, they could provide an email address if they wished to participate in a contest to acquire bol.com gift cards, valued at either 50€, 20€, or 10€. After 7 days of each participant’s first study completion, he or she was contacted again via e-mail and was asked to complete the second part of the study. The email included the link of the second part that redirected participants to the Qualtrics software. After reading a short introduction repeating the supposed purpose of the study and the general instructions, participants had to type in a box the name of their chosen partner. They were then presented with the selected partner’s profile. Afterwards, they answered the satisfaction scale again (Time 2). Following their answering, they were provided with the debriefing form (see Appendix B), which thanked them for their participation in the study, explained the real purpose and asked for their permission for analyzing their personal data. They had to indicate their permission by ticking a box.

Materials

Action versus state-orientation scale. The first independent variable was

decisiveness. Individual differences in decisiveness were measured with the demand-related subscale of Kuhl’s general action control scale (ACS-90; 1994). All items describe specific hypothetical situations and have two alternatives for the participant to select based on how he or she would react. The two alternatives reflect either action or state-orientation. The items display questions that indicate decisiveness or hesitation under demanding conditions (e.g.

“When I know I must finish something soon: A) I have to push myself to get started. B) I find it easy to get it done and over with.).

A reliability analysis with Cronbach’s alpha indicated sufficient validity of the decisiveness’s scale, α= .74.

Dating software. Participants were randomly assigned to the two conditions of the second independent variable “quantity of choice”. In the limited choice condition participants were presented with 6 dating profiles to choose from. In the extensive choice condition participants were presented with 24 profiles. These numbers are consistent with previous studies about choice overload effect (Iyengar & Lepper, 2000; D’Angelo & Toma, 2017). In both conditions, participants were initially informed that they would be shown some profiles and that they had to choose for one dating partner after viewing all available profiles. Only one profile was showing per page and participants had to see all profiles to view the last page that was asking them to type the name of the profile they would like to choose. The software allowed them to move back and forward between profiles by clicking arrows. Profiles included personal information of the daters (name, age, study, personal interests) along with their pictures. Faces of daters were individuals in pictures from the Kandinsky Nijmegen database and pictures provided from Shutterstock. In both conditions the overall profiles were equal in their attractiveness for both males (M = 3.17, SD = .77) and females (M = 3.76, SD = .80), based on independent judges’ ratings on a 7-point likert scale.

Satisfaction scale. The dependent variable, satisfaction with selected partners, was measured both times by the same scale (see Appendix C) that included 6 items about choice satisfaction (e.g. “how satisfied are you with your selected partner?”). All items were measured on a 7-point likert scale with option 1 indicating “not at all” and 7 “extremely” (D’Angelo & Toma, 2017). The scale also included questions to control for confound variables that have been shown to affect satisfaction with partners (D’Angelo & Toma, 2017). These variables are gender, romantic idealization tendencies (“Do you believe in soul mates?”), relationship experience (“How many committed romantic relationships have you had to date?”), online-dating experiences (“Have you ever dated someone you met through online dating?”), behavior toward online-dating (“Would you be willing to use online dating in the future?”), and online-dating efficacy (“I can use online dating to get what I want”).

A reliability analysis with Cronbach’s alpha indicated sufficient validity of the scale of satisfaction questions with high alpha, α= .89.

Results

A regression analysis was calculated to predict satisfaction on Time 2 based on quantity of choice condition, decisiveness scores, their interaction and the covariates, to test our

hypotheses. The result of the regression was not significant, F(8, 72) = .786, p = .616. Only the quantity of choice condition was a significant predictor of satisfaction at time 2, b = 2.764,

t(75) = 2.174, p = .033. Decisiveness scores were not a significant predictor of satisfaction, b =

.148, t(75)= .175, p = .862, neither the interaction of decisiveness and quantity of choice condition, b = -.254, t(75) = -.214, p = .831. Thus, the first four hypotheses were all rejected since decisiveness did not significantly predict satisfaction.

To test whether any effects developed over time a new variable “satisfaction

difference” was computed by subtracting satisfaction levels on Time 1 from satisfaction levels on Time 2. A regression analysis was conducted with satisfaction difference as the dependent variable and the quantity of choice condition, decisiveness, their interaction and the covariates

as predictors. The regression model was non-significant, F(8, 72) = 1.033, p = .420. Quantity of choice condition, b = -.089, t(75) = -.132, p = .895, decisiveness scores, b = .601, t(75) = 1.344, p = .183, and their interaction, b = -.529, t(75) = -.845, p = .401, could not significantly predict the satisfaction difference from initial partner selection to the satisfaction after one week of that selection. The only significant predictor for satisfaction difference was covariate 3 (“Have you ever dated someone you met through online dating?”), b = .841, t(75) = 2.241, p = .028.

Our hypotheses also predicted that participants in the extensive choice condition would report lower satisfaction compared to participants in the limited choice condition. A linear regression analysis was calculated to predict satisfaction at Time 2 based on quantity of choice condition. The regression equation was significant, F(1, 82) = 4.478, p = .037. The quantity of choice condition was an important predictor of satisfaction with partner. For participants in the extensive choice condition the satisfaction levels were significantly higher (M = 4.96, SD = .89) than for participants in the limited choice condition (M = 4.51, SD = .98), failing to support hypothesis 3 (see Figure 1).

23 of the participants delayed completing the second part of the study. To check whether the results are substantially different when these participants are removed, the dataset was analyzed again without them. Thus, those who completed the second part of the study more than 10 days after Time 1 were excluded, leaving the sample with a N of 61. The new sample was consisted of 24 males and 37 females, with a mean age of 23.90 (SD= 2.13). A regression analysis was then conducted with satisfaction at Time 2 as the dependent variable and quantity of choice condition, decisiveness, their interaction and the covariates as the predictors. The regression analysis was non-significant even when excluding those

participants, F(8, 50) = 1.250, p = .291. Only the quantity of choice was a significant predictor for satisfaction with partner, b = 2.738, t(52) = 1.993, p = .052, along with covariates 3 (“Have

you ever dated someone you met through online dating?”), b = -1.404, t(52) = -1. 821, p = .075, and 4 (“Would you be willing to use online dating in the future?”), b = 1.385, t(52)= 1.732, p = .089, that were marginally significant. Decisiveness, b = -.782, t(52) = -.900, p = .373, and the interaction between decisiveness and quantity of choice, b = .426, t(52) = .351, p = .727, were non-significant. Similarly, another regression analysis was computed with

satisfaction difference from Time 1 to Time 2 as the dependent variable and the quantity of choice condition, decisiveness, their interaction and the covariates as the predictors within the new sample. The results of the analysis indicated a non-significant effect of the model, F(8, 50) = .724, p = .670. Quantity of choice condition, b = -.533, t(52) = -.600, p = .551,

decisiveness, b = .537, t(52) = 954, p = .345, and their interaction, b = -.195, t(52) = -.248, p = .805, were all non-significant predictors of satisfaction.

Figure 1. Satisfaction scores across conditions of quantity of choice.

Discussion

The aim of the present study was to investigate if individual differences in decisiveness based on demand-related action versus state-orientation could moderate a choice overload effect in an online-dating situation. Past scientific evidence has indeed indicated that the choice

overload effect can find application to mate selection in online-dating situations with

individuals making poorer choices in extensive choice conditions that in limited ones (Wu & Chiou, 2009; D’Angelo & Toma, 2017). However, there is lack of scientific evidence about the relationship between the choice overload effect and decisiveness. We expected that individual differences in decisiveness could moderate a choice overload experience because several processes underlying the choice overload effect differ between action and

state-oriented individuals. Action compared to state-state-oriented individuals deal more effectively with high-demanding situations by ignoring irrelevant information, increasing the attractiveness of their choices (Beckmann and Kuhl, 1984) and experiencing less regret after actions (McElroy & Dowd, 2007). Thus, we expected them to be more satisfied with their selected partners in a choice overload situation compared to state-oriented participants.

Our analysis did not provide support for a moderating effect of decisiveness on the choice overload effect. Although the quantity of choice was a significant predictor of satisfaction with selected partners, a choice overload effect was not present. Contrary to our hypotheses, participants that had to select a partner among 24 dating profiles, reported higher satisfaction at Time 2 than those that had to select from 6 dating profiles.

However, it is questionable whether a choice overload effect was elicited by our dating website or not. Action and state-oriented individuals show different reactions to negative primes but similar reactions to positive ones (Koole & Fockenberg, 2011) and equal responses to changing demands despite their differences in adjustment (Koole et al., 2012). Furthermore, in low-demanding conditions action-orientation does not predict better self-regulation than state-orientation (Jostmann & Koole, 2006; Koole et al., 2012). Therefore, when they do not perceive a high-demanding choice situation, action and state-oriented individuals could show similar attitudes. A context in which individuals do not experience a choice overload effect, might be a low-demanding situation in which both action and state-oriented individuals can

experience the same choice satisfaction. However, in a high-demanding context that would indeed cause a choice overload effect and dealing with demands’ increase and regret efficiently would be crucial, there are important reasons to believe that action-orientation would moderate the choice overload effect. Therefore, it is important to refer to what should be considered a choice overload in an online-dating situation and thus, discuss the structure of the present dating software.

The pattern of our results, with participants reporting higher satisfaction in the extensive choice condition than in the limited choice condition, might have scientific-based explanations. Previous studies (Wu & Chiu, 2009; D’Angelo & Toma, 2017) have indeed demonstrated that profiles’ size similar to our extensive choice condition (e.g. 24, 40) can decrease satisfaction with the partner and with the enjoyment of the overall selecting process. However, there are also studies indicating comparable levels of satisfaction and enjoyment of the process across conditions (e.g. 4 profiles vs. 64 profiles) (Lenton, Fasolo & Todd, 2008; Lenton & Stewart, 2008). In real dating websites, individuals can browse hundreds or even thousands of profiles. Online daters report spending an average of 5.2 hours/week on searching for profiles, independently of an average of 6.7 hours dedicated to communicating with

potential partners (Finkel et al., 2012). Therefore, the extensive choice condition in the present study may not even be confusing or overwhelming, since 24 profiles is a rather small number for a real online-dating website. Similarly, the reversed effect of satisfaction across conditions may be explained by a tendency of participants in the limited choice condition to seek for more variety. In Iyengar’s and Lepper’s (2000) study with jams, participants were more motivated to buy and more satisfied with their choice in the limited choice condition. However, in the extensive choice condition initial attraction and enjoyment of the task were higher than in the limited choice condition. Choosing for a partner could be a more fun process than choosing for a jam, and therefore individuals might indeed seek for more variety in such cases.

Moreover, a recent meta-analysis by Chernev, Böckenholt and Goodman (2015), proposes some moderators of the choice overload effect. Among those, one is preference certainty. Specifically, the basic feature of this moderator is an “articulated ideal point” that implies well-defined preferences for an item within a specific category. The availability of an ideal point decreases the decision task complexity and thus, the experience of a choice

overload. Furthermore, individuals with such an ideal point tend to prefer and make more purchases from large pools than individuals without it. Therefore, there are greater chances of experiencing a choice overload situation when the preference is uncertain. However, when signing in an online-dating website, online daters already know the characteristics that they are looking for, and they report that they keep browsing a lot of profiles to find partners that match their desired preferences (Finkel et al., 2012). As a result, certainty preference might have affected the choice overload experience in this study. If participants already knew what they prefer in potential partners, many options might have become less after some filtering.

Accordingly, 24 profiles compared to 6 provide more chances to find the a-priori well-defined desired characteristics of a partner. This might be an explanation of the higher satisfaction ratings in the extensive choice condition. Furthermore, as it has already been referred, 24 profiles may not even be a large assortment if online-daters indeed filter options out and keep searching for the ideal partner.

Another explanation of the absence of a choice overload effect of in our study can find its basis on preference reversals (PR) theory. According to this theory, individuals tend to judge differently when they evaluate jointly things than when they evaluate each thing in separation. Namely, they focus on different attributes in the two different modes. For instance, when individuals choose an item among other ones, their selection is usually based on

comparisons with the rest available items of the same category. But when interacting with that item in isolation, then they tend to focus on different attributes that are not stimulated by

comparisons (Hsee, Loewenstein, Blount, & Bazerman, 1999). As a result, they tend to reverse preferences when they switch from a joint to a separate evaluation mode because then they can focus on other features and attributes. Sometimes, these attributes can be more important for their utility than the ones elicited by comparisons. A dating website is usually structured in a way that provides individuals with a side-by-side comparison of available partners. This

represents a joint selection mode that can have similar consequences as the ones that PR theory suggests (Finkel et al., 2012). Our dating website was also structured in a similar way. This pattern might have made participants focused on specific attributes (e.g. attractiveness) but to pay less attention to other attributes (e.g. personality guesses derived by hobbies, studies etc.) that might be more important to make an accurate evaluation of the other person or to conclude if the other person is a good fit for them. Indeed, in speed-dating contexts, participants tend to evaluate easy attributes such as height or weight, more than hard ones and this tendency is usually a predictor of their decisions to go on a date with partners (Lenton & Fransesconi, 2010). This can explain the higher satisfaction ratings in the extensive choice condition since individuals could easily make comparisons. However, individuals might have not evaluated profiles in deep processing and this could explain the absence of the experience of a choice overload effect in the present study. The reasoning behind D’Angelo’s and Toma’s theory of high-stakes context is that online-dating choice affects well-being, emotional involvement and responsibility about justification to significant others. Thus, it can prompt a choice overload effect. However, the single action of just selecting a partner among others in an online-dating website might be a rather weak process to be considered a high-stakes situation that could have such an important impact on the individuals’ choice satisfaction. It is important for future studies to research whether choices and their consequences in online-dating function the same way as in purchasing and consuming environments.

However, it remains questionable why in D’Angelo’s and Toma’s (2017) study a choice overload effect was elicited by using only 24 profiles. There are some differences in design that should be considered when comparing their study with ours. Firstly, they had a limited sample in regards of age. All of their participants were young (M= 20.1, SD = 1.46) compared to the online-daters nowadays that are usually up to 40 years old (D’Angelo & Toma, 2017). In older ages it is hard to find a single partner offline due to limited options in the social milieu while maintaining also a balance between career and personal life

(Valkenburg & Peter, 2007). As a result, older individuals might use online-dating as the main source to find a partner and they might seek for variety. On the contrary, younger people have more chances to meet more single people offline through their social environment and

activities. Having many options both offline and online might be associated with increased chances of experiencing a choice overload effect. Similarly, participants were told that all the other available partners are colleauges from the same university. In general, most dating websites allow users to select the geographical distance as a filter for the profiles they browse. Geographical proximity is important because usually the goal in online-dating is to meet in person with the potential partner (Fiore, Taylor, Zhong, Mendelsohn & Cheshire, 2010). Maybe online-daters exclude some options based on geographical distance. Future studies should investigate what happens under circumstances that proximity is high among the available profiles. If indeed geographical proximity functions as a tool of selection, it could enhance the confusion and thus, the experience of a choice overload effect. In that case, a specific amount of potential partners living close to the user may be perceived differently compared to the same amount of partners living in distant locations. In D’Angelo’s and

Toma’s (2017) study, participants thought that all available partners were studying in the same university as them, and that could have enhanced their choice overload experience. In addition to that, in online-dating there is usually absence of visual features and face-to-face interaction.

Since all the available partners would be in the same university as the participants, they could actively look for them, observe their behavior and make evaluations in person. Such thoughts could enhance the complexity of the process and elicit a choice overload effect. Therefore, new studies are welcome to investigate if age and geographical proximity affect choice strategies and choice overload experiences within this context.

This study had some limitations. Participants were given a cover story which suggested that they are going to test a new dating website. In the cover story, it was not clear if they would indeed go on a date with their selected partners. They knew that the website was primarily for testing purposes. As a result, participants might were aware of the fact that they will not go out with their partners and they are just testing a website with fake profiles. This could have influenced the impact of the process on their decisions, and thus might have affected their responses. Namely, participants might have scored high in satisfaction because they had not experience counterfactual thinking and regret if they realized they are just testing a website. Similarly, they might have completed the scale without much attention or by choosing higher options by chance.

Another limitation is that not all of the pictures that were used for the profiles were very realistic. Most profiles included old-fashioned pictures with white background from previous studies’ databases (see Figure 2), whereas nowadays in dating websites people use more natural pictures of themselves and selfies. This could have negatively affected

participants’ engagement toward the process by making them suspicious of the fakeness of the profiles.

Furthermore, participants were asked to write down on their calendars the day that they had to complete the satisfaction survey again and they were sent a reminder on that specific day. Although this day was a week after their selection for all participants, 43 out of 84 participants delayed to complete the second part of the study more than 10 days. The mean of

the days passed from Time 1 to Time 2 was 9 days (SD = 4.42), with 26 days being the

maximum number of difference. However, even when participants that delayed completing the second part of the study more than 10 days were excluded from the dataset, the pattern of the results remained the same.

Last but not least, individuals that rated profiles’ attractiveness did not report high scores. Although profiles were equal in their attractiveness ratings across the two conditions of quantity of choice, the overall attractiveness ratings of profiles were rather low for both males (M = 3.17, SD = .77) and females (M = 3.76, SD = .80) in a 7-point likert scale with 1

indicating no attractiveness at all and 7 the highest attractiveness. Past studies (Downs & Wright, 1982; Shackelford, Randy & Larsen, 1999) have demonstrated that participants indicate high interjudge agreement when rating others’ attractiveness. This could also have an effect on participants’ experience of a choice overload since they might have filtered some options out based on attractiveness evaluations. Thus, many choices might become just a few after some exclusion.

Future studies should take the above implications into account. Initially, we imply that a dating website or application should have realistic profile numbers across conditions that seem to differ from purchasing contexts. Those numbers should be consistent with real dating websites to increase the probabilities of the choice overload experience. Thus, there is need for more scientific studies to determine how many profiles can predict a choice overload effect in an online-dating situation. The dating website could also include a communication platform, as communication has been shown to affect first impression and evaluation of the potential partner (Finkel et al., 2012). Along with that, more natural photos similar to the pictures that are usually displayed in online-dating websites could be used. It is suggested that such pictures should be high on their attractiveness’ ratings. Finally, when the factors that can cause a choice overload experience in an online-dating situation are determined, the process of selecting a

partner can be perceived as a high demanding situation. In that case, it is suggested that the moderating role of demand-related action versus state-orientation should be explored, as there are important implications of its effects on demanding situations that need to be taken into account. This study has summarized the main individual differences of action and state-oriented individuals that can moderate a demanding situation such as a choice overload effect. As there is no scientific evidence about the moderation of action versus state-orientation on choice overload situations, it could be used as the initial basis of exploring such effects.

References

Beckmann, J., & Kuhl, J. (1984). Altering information to gain action control: Functional aspects of human information processing in decision making. Journal of Research in

Personality, 18(2), 224-237. https://doi.org/10.1016/0092-6566(84)90031-X

D’Angelo, J. D., & Toma, C. L. (2017). There are plenty of fish in the sea: The effects of choice overload and reversibility on online daters’ satisfaction with selected partners. Media Psychology, 20(1), 1-27.

https://doi.org/10.1080/15213269.2015.1121827

Downs, A. C., & Wright, A. D. (1982). Differential conceptions of attractiveness: Subjective and objective ratings. Psychological Reports. 50(1), 282.

http://dx.doi.org/10.2466/pr0.1982.50.1.282

Ellison, N., Heino, R., & Gibbs, J. (2006). Managing impressions online: Self‐presentation processes in the online dating environment. Journal of Computer‐Mediated

Communication, 11(2), 415-441. https://doi.org/10.1111/j.1083-6101.2006.00020.x

Finkel, E. J., Eastwick, P. W., Karney, B. R., Reis, H. T., & Sprecher, S. (2012). Online dating: A critical analysis from the perspective of psychological

science. Psychological Science in the Public Interest, 13(1), 3-66. https://doi.org/10.1177/1529100612436522

Fiore, A. T., Taylor, L. S., Zhong, X., Mendelsohn, G. A., & Cheshire, C. (2010, January). Who's right and who writes: People, profiles, contacts, and replies in online dating. In System Sciences (HICSS), 2010 43rd Hawaii International Conference on (pp. 1-10). IEEE.

Harmon-Jones, E., & Harmon-Jones, C. (2002). Testing the action-based model of cognitive dissonance: The effect of action orientation on postdecisional attitudes. Personality

and Social Psychology Bulletin, 28(6), 711-723.

Haynes, G. A. (2009). Testing the boundaries of the choice overload phenomenon: The effect of number of options and time pressure on decision difficulty and

satisfaction. Psychology & Marketing, 26(3), 204-212. https://doi.org/10.1002/mar.20269

Hsee, C. K., Loewenstein, G. F., Blount, S., & Bazerman, M. H. (1999). Preference reversals between joint and separate evaluations of options: a review and theoretical

analysis. Psychological bulletin, 125(5), 576. doi: 10.1037/0033-2909.125.5.576 Iyengar, S. S., & Lepper, M. R. (2000). When choice is demotivating: Can one desire too

much of a good thing?. Journal of personality and social psychology, 79(6), 995-1006.

Jostmann, N. B., & Koole, S. L. (2006). On the waxing and waning of working memory: Action orientation moderates the impact of demanding relationship primes on

working memory capacity. Personality and Social Psychology Bulletin, 32(12), 1716-1728. https://doi.org/10.1177/0146167206292595

Koole, S. L., & Fockenberg, D. A. (2011). Implicit emotion regulation under demanding conditions: The moderating role of action versus state orientation. Cognition and

Emotion, 25(3), 440-452. https://doi.org/10.1080/02699931.2010.544891

Koole, S. L., Jostmann, N. B., & Baumann, N. (2012). Do Demanding Conditions Help or Hurt Self‐Regulation. Social and Personality Psychology Compass, 6(4), 328-346. https://doi.org/10.1111/j.1751-9004.2012.00425.x

Kuhl, J. (1994). Action versus state orientation: Psychometric properties of the Action Control Scale (ACS-90). In J. Kuhl & J. Beckmann (Eds.), Volition and personality (pp. 47-59). Göttingen, Germany: Hogrefe & Huber.

Kuhl, J., & Beckmann, J. (1994). Volition and personality: State versus action

Lenton, A. P., Fasolo, B., & Todd, P. M. (2008). “Shopping” for a mate: expected versus experienced preferences in online mate choice. IEEE Transactions on Professional

Communication, 51(2), 169-182. doi:10.1109/TPC.2008.2000342

Lenton, A. P., & Francesconi, M. (2010). How humans cognitively manage an abundance of mate options. Psychological Science, 21(4), 528-533.

https://doi.org/10.1177/0956797610364958

Lenton, A. P., & Stewart, A. (2008). Changing her ways: Number of options and mate standard strength impact mate choice strategy and satisfaction. Judgment and

Decision Making, 3, 501–511.

McElroy, T., & Dowd, K. (2007). Action orientation, consistency and feelings of regret.

Judgment and Decision Making, 2(6), 333-341.

McKenna, K. Y., Green, A. S., & Gleason, M. E. (2002). Relationship formation on the Internet: What’s the big attraction?. Journal of social issues, 58(1), 9-31.

https://doi.org/10.1111/1540-4560.00246

Reutskaja, E., & Hogarth, R. M. (2009). Satisfaction in choice as a function of the number of alternatives: When “goods satiate”. Psychology & Marketing, 26(3), 197-203.

https://doi.org/10.1002/mar.20268

Sagi, A., & Friedland, N. (2007). The cost of richness: the effect of the size and diversity of decision sets on post-decision regret. Journal of personality and social

psychology, 93(4), 515-524. http://dx.doi.org/10.1037/0022-3514.93.4.515

Schwartz, B. (2004). The paradox of choice: Why more is less. New York, US: HarperCollins.

Shackelford, T. K., & Larsen, R. J. (1999). Facial attractiveness and physical

health. Evolution and Human Behavior, 20(1), 71-76. https://doi.org/10.1016/S1090-5138(98)00036-1

Valkenburg, P. M., & Peter, J. (2007). Who visits online dating sites? Exploring some characteristics of online daters. CyberPsychology & Behavior, 10(6), 849-852. https://doi.org/10.1089/cpb.2007.9941

Weiten, W., Dunn, D. S., & Hammer, E. Y. (2014). Psychology applied to modern life:

Appendices Appendix A Cover Story Welkom!

Dit onderzoek bestaat uit twee delen:

1. Het testen van een nieuw online dating systeem. Voor deze test vragen we je nu deel te nemen (15 minuten), en nog een keer over een week (5-10 minuten). 2. Een studie over daadkracht. Voor deze studie zul je een vragenlijst invullen direct

na het testen van de dating site vandaag (5 minuten).

Voordat de onderzoeken van start gaan, willen we je vragen om de informatie op de volgende pagina goed door te lezen. Indien je aan de studies wilt meedoen kun je het hokje onderaan de pagina aanvinken.

Wij zijn bezig met het ontwerpen van een online dating systeem dat wordt ontwikkeld voor Nederlandse studenten. Wij willen dan ook jongvolwassenen (20-28 jaar) werven om de bètaversie van het systeem te testen. Je zal een selectie van profielen te zien krijgen met foto’s en beschrijvingen van leden van de dating website. Nadat je de profielen bekeken hebt, zullen we je vragen om een potentiële date partner te selecteren. Vervolgens zal je gevraagd worden om het online dating systeem een cijfer en feedback te geven. Na een week zal er nog eens naar je mening over het systeem gevraagd worden. De feedback zal gebruikt worden om het systeem te verbeteren voor de officiële lancering.

Voor de studie over daadkracht wordt je gevraagd te kiezen uit 12 gedragingen die jou het beste beschrijven.

Dit onderzoek is uitsluitend gericht op jouw mening. Er zijn dus geen goede of foute antwoorden. Tevens is deelname aan dit onderzoek geanonimiseerd en worden de gegevens strikt vertrouwelijk verwerkt en bewaard. Er wordt je gevraagd naar een email adres om de

gegevens te kunnen koppelen met de gegevens van de tweede vragenlijst. Direct na het invullen van de tweede studie worden jouw ingevulde antwoorden ontkoppelt van het ingevulde email adres om anonimiteit te waarborgen.

Deelname aan dit onderzoek is geheel vrijwillig. Je kunt op elk moment stoppen als je dat wilt.

Als dank voor je deelname, verloten we onder alle deelnemers 2 bol.com bonnen ter waarde van 50 euro en 2 bol.com bonnen ter waarde van 25 euro.

Het onderzoek wordt uitgevoerd onder coördinatie van dr. Lottie Bullens, Universiteit Leiden (l.bullens@fsw.leidenuniv.nl). Voor vragen, klachten of andere informatie kun je bij haar terecht.

Ik heb de informatiebrief voor de proefpersoon gelezen. Ik had voldoende tijd om te beslissen of ik meedoe. Ik weet dat meedoen helemaal vrijwillig is. Ik weet dat ik op ieder moment kan beslissen om toch niet mee te doen of te stoppen. Daarvoor hoef ik geen reden te geven. Mijn antwoorden worden anoniem/ gecodeerd verwerkt. Ik geef toestemming om mijn gegevens te gebruiken, voor de doelen die in de informatiebrief staan.

Appendix B Debriefing Form Bedankt voor je deelname aan dit onderzoek!

Het doel van dit onderzoek was om te bestuderen of de hoeveelheid keuzemogelijkheden de tevredenheid met een keuze beïnvloedt in een dating context. Daarom kreeg de helft van de deelnemers 6 profielen te zien en de andere helft andere 24.

Wij verwachten dat minder daadkrachtige mensen tevredener zullen zijn met hun keuze als zij kiezen uit 24 profielen dan als zij kiezen uit 6 profielen. Daarnaast verwachten we dat mensen die daadkrachtiger zijn, tevredener zijn met hun keuze als zij kiezen uit 24 profielen dan mensen die minder daadkrachtig zijn.

Jouw bijdrage aan dit onderzoek was belangrijk, omdat de effecten van een overvloed aan keuzes veelvuldig is onderzocht binnen de consumentenpsychologie, maar nog onvoldoende in situaties zoals online daten.

Om de studie zo realistisch mogelijk te maken, hebben we je geïnformeerd dat we een bètaversie van een nieuwe online dating website voor studenten aan het onderzoeken zijn. In werkelijkheid zijn de profielen en de website alleen ontwikkeld voor de huidige studie, en zal deze niet online gelanceerd worden.

Iedere verwachting die onze deelnemers zouden hebben over hun niveau van daadkracht en het mogelijke effect daarvan op tevredenheid met hun keuze, had onze resultaten kunnen vertekenen. Om verwachtingseffecten te vermijden hebben we je geïnformeerd dat de daadkrachtstudie niet gerelateerd was aan het testen van de website. Om onze hypotheses te kunnen testen, zouden we echter graag de resultaten van de daadkracht vragenlijst relateren aan de tevredenheid met de gekozen date partner. De informatie die je hebt verschaft, zal uiteraard anoniem blijven tijdens dit proces. Zoals je weet, is deelname aan deze studie vrijwillig. Als je wil, mag je je terugtrekken uit de studie na het lezen van deze informatie en

zullen al je gegevens vernietigd worden. Dit heeft geen enkele gevolgen voor je kansen in de loterij.

Mocht je nog vragen hebben over de huidige studie, e-mail dan gerust dr. Lottie Bullens, l.bullens@fsw.leidenuniv.nl

o Ik geef toestemming voor het analyseren van mijn anonieme data voor het doel van de huidige studie.

o Ik geef geen toestemming voor het analyseren van mijn anonieme data voor het doel van de huidige studie en wil dat mijn data vernietigd worden.

Appendix C Satisfaction Scale 1. In hoeverre ben je tevreden met je gekozen profiel?

2. In hoeverre lijkt de persoon van wie jij het profiel hebt geselecteerd jou leuk? 3. In hoeverre zou je deze persoon graag in het echt ontmoeten?

4. In hoeverre vind je je gekozen profiel aantrekkelijk?

5. In hoeverre was je blij met het aanbod van de profielen dat je had? 6. In hoeverre vond je de overige opties aantrekkelijk?