ArticlePDF Available

A Rejection Mind-Set: Choice Overload in Online Dating


Abstract and Figures

The paradox of modern dating is that online platforms provide more opportunities to find a romantic partner than ever before, but people are nevertheless more likely to be single. We hypothesized the existence of a rejection mind-set: The continued access to virtually unlimited potential partners makes people more pessimistic and rejecting. Across three studies, participants immediately started to reject more hypothetical and actual partners when dating online, cumulating on average in a decrease of 27% in chance on acceptance from the first to the last partner option. This was explained by an overall decline in satisfaction with pictures and perceived dating success. For women, the rejection mind-set also resulted in a decreasing likelihood of having romantic matches. Our findings suggest that people gradually “close off” from mating opportunities when online dating.
Content may be subject to copyright.
A Rejection Mind-Set: Choice Overload
in Online Dating
Tila M. Pronk
and Jaap J. A. Denissen
The paradox of modern dating is that online platforms provide more opportunities to find a romantic partner than ever before,
but people are nevertheless more likely to be single. We hypothesized the existence of a rejection mind-set: The continued access
to virtually unlimited potential partners makes people more pessimistic and rejecting. Across three studies, participants imme-
diately started to reject more hypothetical and actual partners when dating online, cumulating on average in a decrease of 27% in
chance on acceptance from the first to the last partner option. This was explained by an overall decline in satisfaction with pictures
and perceived dating success. For women, the rejection mind-set also resulted in a decreasing likelihood of having romantic
matches. Our findings suggest that people gradually “close off” from mating opportunities when online dating.
romantic partner selection, online dating, mate choice, choice overload, interpersonal relationships
The dating landscape has changed drastically over the past
decade, with more and more people looking for a partner online
(Hobbs, Owen, & Gerber, 2017). People have never been able to
select partners among such an enormous pool of options. As an
example, the 10 million active daily users of the popular online
dating application Tinder are on average presented with 140
partner options a day (Smith, 2018). While one may expect this
drastic increase in mating opportunities to result in an increasing
number of romantic relationships, the opposite has occurred: The
rise of online dating coincided with an increase in the amount of
singles in society (Centraal Bureau voor de Statistiek, 2019;
Copen, Daniels, Vespa, & Mosher, 2012; DePaulo, 2017). What
could explain this paradox in modern dating?
The abundance of choice in online dating is one of the key
factors which explains its success (Lenton & Stewart, 2008).
People like having many options to choose from, and the like-
lihood of finding an option that matches someone’s individual
preference should logically increase with more choice (Lancas-
ter, 1990; Patall, Cooper, & Robinson, 2008). However, having
extensive choice can have various adverse effects, such as
paralysis (i.e., not making any decision at all) and decreased
satisfaction (Iyengar & Lepper, 2000; Scheibehenne, Greifene-
der, & Todd, 2010; Schwartz, 2004). In fact, it seems that peo-
ple generally experience less benefits when they have more
choice. This observation is reminiscent of the basic economic
principle of diminishing returns (Brue, 1993; Shephard & Fa¨re,
1974), in which each unit that is sequentially added to the pro-
duction process results in less profits.
There is some indirect evidence that having more choice in
the domain of dating also has negative consequences. For
example, when asked to pick the best partner, access to more
partner profiles resulted in more searching, more time spent
on evaluating bad choice options, and a lower likelihood of
selecting the option with the best personal fit (Wu & Chiou,
2009). Likewise, when a choice set increases, people end up
being less satisfied with their ultimate partner choice and more
prone to reverse their decision (D’Angelo & Toma, 2017). The
adverse effects of choice overload are also mentioned in arti-
cles in popular media mentioning phenomena such as “Tinder
fatigue” (Beck, 2016) or “dating burnout” (Blair, 2017).
To shed more light on the paradoxical effects of modern dat-
ing, we studied what happens once people enter an online dat-
ing environment. Our innovative design allowed us to observe
how people’s partner choices unfold when people are presented
with partner options sequentially—as opposed to simultane-
ously (D’Angelo & Toma, 2017; Wu & Chiou, 2009). Our
main expectation was that online dating will set off a rejection
mind-set, leading people to become increasingly likely to reject
partners to the extent that they have been presented with more
options. Secondly, we explored the question of timing: How
quickly will the rejection mind-set kick in? We did not have
any a priori hypothesis on what an ideal choice set would be but
instead explored a potential “break point” in the tendency to
Tilburg University, Tilburg, the Netherlands
Corresponding Author:
Tila M. Pronk, Tilburg University, Prof. Cobbenhagenlaan 225, 5037 DB Tilburg,
the Netherlands.
Social Psychological and
Personality Science
ªThe Author(s) 2019
Article reuse guidelines:
DOI: 10.1177/1948550619866189
reject. Third, we tested which psychological processes may
account for a change in mating decisions.
The Present Research
We tested the existence of a rejection mind-set in online dating
across three studies. In Study 1, we presented people with pic-
tures of hypothetical partners, to test if and when people’s gen-
eral choice behavior would change. In Study 2, we presented
people with pictures of partners that were actually available
and tested the gradual development of their choice behaviors
as well as their success rate in terms of mutual interest (i.e.,
matches). In Study 3, we explored potential underlying psycho-
logical mechanisms. Specifically, and in line with choice over-
load literature, we explored whether the rejection mind-set may
be due to people experiencing lower choice satisfaction and
less success over the course of online dating. As an additional
goal, we explored the potential moderating role of gender. In all
old—a group that makes up 79%of all users of online dating
applications (Smith, 2018).
All studies described below received approval from the ethi-
cal review board. We uploaded the working data files and
R scripts for analyzing the data of all studies on the Open Sci-
ence Framework ( We computed post hoc
power analyses via the SIMR package, Version 1.0.3 (Green &
MacLeod, 2016). This analysis indicated that we had 100%,
92%,and100%power to confirm the statistical significance
(a¼.05) of a logistic regression coefficient of b¼.10 in
Studies 1, 2, and 3, respectively. Such a coefficient corresponds
to a 9.5%decrease in the odds of accepting a partner after one
standard deviation (SD) increase in our focal independent vari-
able (see below).
Study 1
Study 1 provided a first test of our main hypothesis. Previous
research showed that a set of potential partners ideally consist
of 20–50 options (Lenton, Fasolo, & Todd, 2008), and we
expected that changes in acceptance may occur when a set goes
beyond this range. We therefore randomly divided participants
into two conditions, in which they were either presented with
45 partner options (within the ideal range) or with 90 partner
options (double the ideal range). We aimed to test whether
acceptance rate (i.e., the chance of accepting each consecutive
potential partner) would decrease over the course of online dat-
ing, and whether this effect differed depending on condition
and gender.
Participants and Design
Participants were recruited via Amazon Mechanical Turk
(Buhrmester, Kwang, & Gosling, 2011), with the following
information: “In this survey, you will be rating pictures of
potential romantic partners. This study is ONLY available for
participants between 18 and 30 years old, who are heterosex-
ual & single.” Participants received US$2 for taking part in
the study.
A total of 423 individuals participated. We deleted 108 parti-
cipants from our data set because they were not single (N¼94),
outside the appropriate age range (N¼6), not heterosexual (N¼
1), or with missing data on key variables (N¼7). The remaining
data set of 315 participants consisted of an approximately equal
amount of men (N¼159) and women (N¼156), in the age
rangefrom18to30yearsold(M¼26.07, SD ¼2.94).
Procedure and Materials
Participants filled out the questionnaire programmed in Qual-
trics (version December 2016). After reading general informa-
tion about the study and giving consent for participation,
participants started with the online dating task, which was mod-
eled after the dating application Tinder. In this task, partici-
pants were exposed to either 45 or 90 pictures of
hypothetical potential partners. These pictures were selected
after extensive pretesting on characteristics that we deemed
important: perceived age, level of attractiveness, and appropri-
ateness for use as online dating pictures. The final set contained
pictures of hypothetical potential partners that were perceived
to be between 18 and 30 years old (matching the age range
of our participants), appropriate as online dating pictures, and
slightly above average in attractiveness level (for a detailed
report, see
After given consent for participation, participants received
the following instructions: “In the following task, 45/90
[depending on condition] pictures of potential partners will
be presented on the screen. We kindly ask you to respond to
these pictures, by clicking the green heart to accept, or the red
cross to reject the picture.” The pictures appeared in random
order in the middle of the screen one by one. We counted the
number of pictures that had been presented previously and
saved this information as a sequence variable (i.e., a continuous
variable having a value of 9 for the 10th picture, 10 for the 11th
picture, etc.). There was no time limit, and a new picture was
presented immediately after participants gave a response on the
previous picture.
After the online dating task, participants filled out several
questionnaires (for a complete list of all the variables that were
assessed, see, including a question on par-
ticipants’ level of satisfaction (a measure we will discuss under
the heading “Additional Analyses across Studies”): “Are you
satisfied with the choices you made for the people that you
have accepted?” on a scale from 1 (not at all)to10(very
much). Thereafter, participants were informed about the main
goal of the study, thanked for their participation, and paid.
Analytic Strategy
In all studies, we used R Version 3.4.1 (R Core Team, 2013)
and lme4 Version 1.1.13 (Bates, Ma¨chler, Bolker, & Walker,
2014) to model the relationship between sequence, condition
2Social Psychological and Personality Science XX(X)
(0 ¼90 pictures,1¼45 pictures), gender (0 ¼male,1¼
female), and the acceptance level of pictures of potential part-
ners. For each study, we also ran a model in which the effect of
sequence interacted with gender. Because people likely differ
in selectivity, we applied random effects modeling, with
choices nested within participants. In all analyses, we modeled
random intercepts only (and not random slopes). For models
with binomial outcomes (e.g., choices, matches), we computed
logistic regressions. pValues of all multilevel coefficients were
computed with the package lmerTest, Version 2.0-33 (Kuznet-
sova, Brockhoff, & Christensen, 2017).
Descriptive Statistics
Means and SDs of all relevant variables across all three studies
are displayed in Table 1.
Main Effect of Condition, Gender, and Sequence
We first entered condition, sequence, and gender (not the inter-
action terms) into the model as fixed effects and the intercept as
random effect (to account for between-person differences in
selectivity). Our results showed that condition did not signifi-
cantly affect acceptance rate, b¼0.038, p¼.80, 95%confi-
dence interval (CI) [0.260, 0.336]. Gender did affect
acceptance rate, b¼1.94, p< .001, 95%CI [2.240,
1.645], with men accepting on average 34%more pictures
of potential partners compared to women. Sequence also
affected acceptance rate, b¼0.123, p< .001, 95%CI
[0.159, 0.087]. As compared to the first picture, the chance
on acceptance on average decreased with 29%over the task.
Given that both Studies 1 and 3 used the same set of pictures,
we represented the effect of sequence on acceptance rate of
these studies combined in Figure 1.
Interaction Effects of Sequence With Condition
and Gender
We also ran the model with sequence, condition, and gender as
well as their two- and three-way interactions added as fixed
effects. Results indicated that the two-way interaction between
sequence and condition (b¼0.231, p¼.006, 95%CI
[0.394, 0.068]) and the three-way interaction between
sequence, condition, and gender (b¼0.278, p¼.012, 95%
CI [0.061, 0.496]) were significant. Subsequent analyses on the
simple slopes revealed that the effect of sequence was signifi-
cant for all groups (p< .006), except for women presented with
45 pictures (b¼0.095, p¼.16, 95%CI [0.227, 0.037]).
Plotting the effects indicated that only the effects of sequence
and gender appeared consistent, which is why we did not fur-
ther investigate the effect of condition.
Existence of Break Points
We used the strucchange package Version 1.5-1 in R (Zeileis,
Leisch, Hornik, & Kleiber, 2002) to identify the so-called break
points: moments in the study during which responses started to
shift. Across genders, the decrease in acceptance was steep
between Photos 1 and 13 and became relatively flat thereafter.
Study 1 showed support for our main hypothesis that people
become increasingly likely to reject potential partners while
online dating—a trend that was especially prevalent in the first
dozen pictures.
Study 2
Our main aim of Study 2 was to replicate the effect of sequence
on partner acceptance with consequential partner choices.
Additionally, we wanted to test whether the likelihood of hav-
ing a “match” (a case of mutual acceptance: participant A
accepts the picture of participant B and vice versa) is also
affected by sequence. We preregistered Study 2 (https://aspre but slightly deviated from the preregis-
tration in the execution of the study. Specifically, due to time
constraints, we did not assess whether level of satisfaction with
the pictures and perceived success changed over time. Instead,
we added these variables to Study 3.
Participants were recruited by handing out flyers at the univer-
sity campus, via e-mail, and social media channels. We invited
single, heterosexual people between 18 and 30 years to partic-
ipate in a study on online dating. As compensation, participants
could enter a raffle to win one of four cinema tickets. Sample
size was determined by the maximum number of participants
Table 1. Means and Standard Deviations of All Relevant Variables
Across Studies 1–3.
Variable Mean SD
Study 1 (45/90 pictures, depending on condition)
Average acceptance 23.94/45.55 11.64/23.94
Study 2 (40/45 pictures, depending on gender)
Average acceptance 12.14 8.17
Average match 2.15 2.48
Study 3 (50 pictures)
Average acceptance 26.07 12.4
Perceived success 1 4.01 2.47
Perceived success 2 2.89 2.26
Satisfaction with pictures 70.56 25.21
Note. Perceived success 1: “How many of the people from the previous block
do you think would have accepted your picture?” (1–10); Perceived success 2:
“How many matches do you think would you have in the previous block?”
(1–10); Satisfaction with pictures: “I am satisfied with the quality of the pictures
in the previous block” (1–100). All 3 items in Study 3 were logarithmized to
reduce skew.
Pronk and Denissen 3
we could recruit in the weeks before the study started, while
keeping the gender ratio equal.
A total of 170 people signed up for the study by sending in a
picture, and 164 individuals continued by participating in the
online dating task. We deleted six participants from our data set
because they were not single (N¼5) or because they requested
their data to be removed after participation (N¼1). The remain-
ing data set of 158 participants consisted of 82 men and 76
women between 18 and 29 years old (M¼22.24, SD ¼2.48).
Procedure and Materials
We instructed participants to send in a picture in color, in which
their face was clearly visible and there were no other people
displayed. We created two sets of groups, each comprising of
45 men and 40 women. Participants received a link to the ques-
tionnaire via e-mail, that startedbyarequestforconsentfor
participation. The online dating task started with similar
instructions as in Study 1, with the following addition: “The
people in the pictures also rate your picture, and you can really
get a ‘match.’” Participants were not informed on the set size
prior to the task.
After the online dating task, participants filled out several
questionnaires (see, including the item
assessing level of satisfaction (see Study 1). Participants were
then informed on the goal of the study, and thanked. Within
1 week after completion of the study, participants received
an e-mail containing the picture and the e-mail address of
their matches, and information on whether they had won a
cinema ticket.
Analytic Strategy
We used the same analytic strategy as described in Study 1 but
nested within targets to account for differences in attractiveness.
We modeled the relationship between sequence, gender, and the
acceptance level of pictures of potential partners, see Figure 2.
Figure 1. The effect of sequence on choice behavior for women and men in Studies 1 and 3. Higher scores represent a higher chance on
acceptance of the picture of a potential partner. The lines in the figure represent smoothed predicted means, using the “loess” algorithm of the
ggplot2 package (version 2.2.1). The gray area around the lines represent the 95% confidence intervals around these predicted means.
4Social Psychological and Personality Science XX(X)
Main and Interactive Effect of Sequence and Gender
on Acceptance Level
We first entered sequence and gender into the model as fixed
effects and the intercept for subjects as random effect. Our
results showed that gender affected acceptance rate, b¼
1.897, p< .001, 95%CI [2.617, 1.197], with men accept-
ingonaverage25%more potential partners compared to
women. Sequence also affected acceptance rate, b¼0.130,
p< .001, 95%CI [0.205, 0.056], lowering it by about
29%over the entire task. The interaction between sequence and
gender did not significantly affect acceptance rate, b¼0.034,
p¼.65, 95%CI [0.183, 0.115]. A break point occurred at the
31st photo for women and at the 34th photo for men. The
decrease in acceptance was relatively flat up to this break point
and became steeper thereafter.
Main and Interactive Effect of Sequence and Gender
on Match Rate
We modeled the relationship between sequence, gender, and
match probability using multilevel modeling in R (specifying a
binary outcome variable, with 0 ¼no match vs. 1 ¼match).
Sequence, gender, and their interaction were entered into the
model as fixed effects and the intercept for subjects as random
effect. Our results showed that neither the effect of gender
(b¼0.097, p¼.74, 95%CI [0.483, 0.681]) nor the effect
of sequence (b¼0.172, p¼.08, 95%CI [0.018, 0.364]) was
statistically significant. However, the interaction between
sequence and gender on match rate was significant, b¼
0.379, p¼.003, 95%CI [0.628, 0.131], see Figure 3 (for
an alternative approach of this analysis,see
Simple slope analyses showed that women’s chances on a match
decreased over the task, b¼0.206, p¼.011, 95%CI [0.366,
0.048], lowering it by a total of 70%.Thiseffectwasnotsig-
nificant for men, b¼0.172, p¼.078, 95%CI [0.018, 0.365].
Only for women, a break point occurred at the 8th picture, so
that there was a steep decrease in match probability during
these first trials, after which the effect flattened out.
In Study 2, we replicated the findings of Study 1, demonstrat-
ing that people become more likely to reject—actually avail-
able—potential partners. For women, the likelihood of
having a match also declined.
As compared to Study 1, overall rejection rate in Study 2
was higher and the break point occurred much later. This may
be due to some key differences between the two studies. First,
Study 2 involved real-life mating decisions. Second, we
recruited Amazon Mechanical Turk workers in Study 1 and
Dutch university students in Study 2. Third, we used a prese-
lected set of above-average attractive partners in Study 1 versus
a set of actually available partners with various ranges of attrac-
tiveness in Study 2.
Study 3
In Study 3, we aimed to replicate the rejection mind-set effect
once more and explored potential underlying psychological
mechanisms (i.e., level of satisfaction with pictures and percep-
tion of own dating success). To do so, we added questions
about participants’ experience with the online dating task after
every block of 10 pictures.
Figure 2. The effect of sequence on choice behavior for women and
men in Study 2. Higher scores represent a higher chance on accep-
tance of the picture. The lines in the figure represent smoothed pre-
dicted means, using the “loess” algorithm of the ggplot2 package. The
gray area around the lines represent the 95% confidence intervals
around these predicted means.
Figure 3. The effect of sequence on match rate for women and men
in Study 2. Higher scores represent a higher chance on having a match
with a potential partner. The lines in the figure represent smoothed
predicted means, using the “loess” algorithm of the ggplot2 package.
The gray area around the lines represent the 95% confidence intervals
around these predicted means.
Pronk and Denissen 5
We invited single, heterosexual people between 18 and 30
years to participate on M-Turk for US$2. As in Study 1, we
aimed to recruit 400 participants.
A total of 402 individuals participated. We deleted 93 parti-
cipants from our data set because they were not single (N¼
90), outside the appropriate age range (N¼3), or with missing
data on key variables (N¼4). The remaining data set of 305
participants consisted of an approximately equal amount of
men (N¼150) and women (N¼155) from 18 to 30 years old
(M¼26.16, SD ¼2.80).
Procedure and Materials
After giving consent for participation, participants completed
the same online dating task as described in Study 1. All parti-
cipants were exposed to 50 pictures of hypothetical partners
(derived from the same picture set we used in Study 1), which
were divided in 5 blocks each containing 10 pictures. Partici-
pants were instructed beforehand about the set size and the
division of pictures into blocks of 10. We distributed the pic-
tures so that the mean attractiveness level of the pictures in
each block was similar, and we counterbalanced the presenta-
tion of the blocks between participants (see
zntb6/). The sequence variable thus referred to the order of the
blocks in this analysis. Within each block, the pictures were
randomly presented.
In between blocks, participants answered several questions
about their experience with the task in the past block. We
explored the effect of different mediators (for a complete list,
see, including one question about level
of satisfaction with the quality of the pictures (“I am satisfied
with the quality of the pictures in the previous block”), mea-
sured with a slider (0–100; the default starting point of 50).
We also included two questions on participants’ perception
of their own success (“How many of the people from the pre-
vious block do you think would have accepted your picture?”
and “How many matches do you think would you have in the
previous block?”), for which participants entered a number
between 0 and 10. Because of the high correlation between
these 2 items (rbetween .65 and .75 across blocks), we com-
puted the mean score as our indicator of perceived dating
At the end of the final block, participants filled out several
questionnaires (see, including the item
assessing their level of satisfaction (see Study 1). Then, parti-
cipants were informed on the main goal of the study, thanked
for their participation, and paid.
Analytic Strategy
We used the same analytic strategy as described in Study 1.
Main and Interactive Effect of Sequence and Gender
We visualized the effects of (block) sequence and gender on
acceptance rate in Figure 1. We first entered sequence and gen-
der into the model as fixed effects and the intercept for subjects
as random effect. Our results showed that gender affected
acceptance rate, b¼1.019, p< .001, 95%CI [1.304,
0.736], with men accepting on average 19%more potential
partners compared to women. Sequence did not affect accep-
tance rate, b¼0.035, p¼0.1917, 95%CI [0.088,
0.0176]. The interaction between sequence and gender signifi-
cantly affected acceptance rate, b¼0.083, p¼.03, 95%CI
[0.155, 0.010]. Simple slope analyses revealed that the
effect of sequence was significant for women (decreasing with
29%overtheentiretask),b¼0.117, p< .001, 95%CI
[0.1668, 0.0674], but not for men, b¼0.035, p¼0.19,
95%CI [0.088, 0.0175]. Only for women, a break point
occurred at the 16th picture, so that the decrease in acceptance
was steep up to this break point and became relatively flat
Mediation Effect of Psychological Constructs
We performed all mediation analyses using lavaan Version 0.5-
23.1097 (Rosseel, 2012). We specified a model with a direct
effect of sequence on acceptance as well as an indirect effect
consisting of the effect of sequence on the mediator, multiplied
by the effect of the mediator on acceptance. Mediation was
established by testing this indirect effect. We used the lavaan.-
survey package, Version (Oberski, 2014) to account for
the nested structure of our data.
Results of the mediation analysis are depicted in Figure 4.
As can be seen, two factors turned out to be significant media-
tors: level of satisfaction with pictures and perception of own
dating success. These items were logarithmized to reduce skew
(right skew in perceived success, left skew in satisfaction).
When we included both mediating paths simultaneously, there
was a significant indirect effect of sequence on acceptance via
reduced satisfaction with pictures, b¼0.004, p¼.006, 95%
CI [0.007, 0.001]. Similarly, there was a significant indirect
effect via reduced perceptions of own dating success, b¼
0.004, p¼.002, 95%CI [0.006, 0.001]. After controlling
for both independent indirect effects, the direct effect was no
longer significant, b¼0.002, p¼.53, 95%CI [0.007,
0.004], showing full mediation. Constraining regression paths
to be equal across gender did not result in a decrease in fit when
compared to a model with unconstrained regression paths,
(5) ¼7.98, p¼.16 (for additional analyses on the media-
tors, see
The results of Study 3 again showed that women (but not men)
became more likely to reject partner options when online dat-
ing. Moreover, Study 3 revealed that people experienced a
6Social Psychological and Personality Science XX(X)
decrease in satisfaction with the pictures and a decrease in per-
ceptions of their own dating success, which in turn accounted
for the increasing tendency to reject potential partners.
Additional Analyses Across All Studies
Relationship Between Tendency to Reject and
One could argue that an increased rejection rate may not neces-
sarily be harmful; it may instead help people focus on the best
possible partner options and thus leave them more satisfied
with the set of partners they accepted. However, across all three
studies, there was a negative relationship between overall
rejection behavior and satisfaction with the partner options par-
ticipants accepted (r¼.34 in Study 1, .20 in Study 2, and
.22 in Study 3, p.01). Thus, people who were more reject-
ing were more likely to be less satisfied with the smaller num-
ber of partner options they did accept compared to people who
were overall more accepting.
Effect of Sequence Versus Cumulative Rejection
We assumed that the increasing tendency to reject is due to
sequence (i.e., more choice options), but it may also be
explained by cumulative rejection (i.e., more previous rejec-
tions)— variables that were highly (but not perfectly) corre-
lated in our studies (*r¼.70). To disentangle these effects,
we included a “rejection count” variable in our models together
with the sequence variable as well as a between-person variable
representing a participant’s average likelihood to reject part-
ners. Results indicated that it was the cumulative rejection of
profiles that increased the likelihood of rejecting a following
partner (Study 1: b¼0.749, p< .001, 95%CI [0.890,
0.610]; Study 2: b¼1.359, p< .001, 95%CI [1.708,
1.016]; Study 3: b¼0.884, p< .001, 95%CI [1.024,
0.745]). After controlling for this tendency, the sequence
effect actually became positive (Study 1: b¼0.340, p<
.001, 95%CI [0.248, 0.434]; Study 2: b¼0.815, p< .001,
95%CI [0.574, 1.059]; Study 3: b¼0.525, p< .001, 95%CI
[0.422, 0.630])—an effect that should be interpreted with cau-
tion given concerns of multicollinearity.
General Discussion
The findings of all three studies showed support for our main
hypothesis that a higher number of partner options sets off a
rejection mind-set: People become increasingly likely to reject
potential partners to the extent that they are presented with
more options. In Studies 1 and 3, the first partner option had the
highest chance of being accepted, after which the acceptance
rate decreased up to stabilization after circa a dozen choices.
choices, after which decrease accelerated—suggesting that the
rejection mind-set might be postponed when participants
expect real interactions.
In all studies, women became increasingly likely to reject
potential partners, while for men this effect was either weaker
(Study 1), similar (Study 2), or not significant (Study 3). In
Study 2, the likelihood of finding a match only significantly
decreased for women. Overall, the adverse effects of choice
abundance in dating thus seem to apply particularly to
women—the gender that is already much less likely to accept
potential partners to begin with, possibly consistent with evolu-
tionary pressures (Buss & Schmitt, 1993). As a consequence,
the initial benefit women have in their likelihood of having a
match dissolved in the process of online dating.
The results of Study 3 suggested two underlying psycho-
logical mechanisms of the rejection mind-set: increased dis-
satisfaction with the pictures and increased pessimism about
one’s chances of finding a partner through this platform.
These findings are seemingly in line with choice overload
theory (Iyengar & Lepper, 2000; Scheibehenne et al.,
2010; Schwartz, 2004). However, results of our cumulating
rejection variable showed that more choice does not always
lead to more rejection. Instead, participants became increas-
ingly rejecting depending on how many partners they
already rejected. This may explain why we did not find
an effect for condition in Study 1, and it might also be the
reason why the sequence effect was sometimes stronger for
women: Because women typically reject more, they might
also cumulate rejections more quickly and thus more easily
adopt a rejection mind-set. Futureresearchusinganexperi-
mental approach is needed to disentangle the effect of rejec-
tion behavior from cumulative choice.
We also aimed to gain some insight into the question of
timing: When does the rejection mind-set kick in? Current
findings do not give a clear answer to this question, given that
results of the exploratory break-point analyses were relatively
inconsistent. However, across all studies, acceptance rate
decreased over the course of online dating. An implication
of this main finding is that people may benefit from restricting
their search (or at least the amount of rejections during this
search) when online dating. One may even consider deciding
on one potential partner at a time, followed by a substantial
period of time to “de-habituate.” Another implication is that
the order in which someone’s picture is presented has conse-
quences on this person’s chance on romantic success. In
line with this observation, Tinder now offers users the
Figure 4. The standardized regression coefficients for the relation-
ship between block sequence and acceptance rate, as mediated by
satisfaction with the pictures and participants’ perception of their
romantic success in each block. *p< .01. **p< .001.
Pronk and Denissen 7
option to pay to have their picture shown first in the row
(Robinson, 2018).
Dating is not the only domain in life in which choice options
have vastly expanded. From relatively mundane daily choices
(e.g., grocery shopping) to major life decisions (e.g., buying
a house), people now face more options than ever before. It
remains to be tested whether a rejection mind-set also applies
to these contexts. Also, it would be interesting to test whether
the rejection mind-set is specific for online dating or whether it
generalizes to other forms of dating (e.g., speed dating).
Historically speaking, people’s chances of finding a suitable
mate have been limited, leading to a strong preference for dat-
ing environments with more partner options. Online dating has
smartly catered to this appetite for choice, offering users access
to a virtually endless pool of partner options. Our research
reveals that—instead of benefiting from more choice options,
and coming closer to finding out what the best possible partner
is—the stream of partner profiles can set in motion an overall
feeling of dissatisfaction and pessimism about finding a mate,
which leads users to gradually “close off” from mating oppor-
tunities. Our findings might therefore explain why people are
increasingly dissatisfied and frustrated by modern dating.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to
the research, authorship, and/or publication of this article.
The author(s) received no financial support for the research, author-
ship, and/or publication of this article.
Bates, D., Ma¨chler, M., Bolker, B., & Walker, S. (2014). Fitting linear
mixed-effects models using lme4. ArXiv:1406.5823 [Stat] .
Retrieved from
Beck, J. (2016, October 25). The rise of dating-app fatigue. Retrieved
September 18, 2018, from The Atlantic website: https://www.thea
Blair, O. (2017). There’s one very big problem with Tinder. Retrieved
September 18, 2018, from The Independent website: http://www.
Journal of Economic Perspectives,7, 185–192.
Buhrmester, M., Kwang, T., & Gosling, S. D. (2011). Amazon’s
Mechanical Turk: A new source of inexpensive, yet high-quality,
data? Perspectives on Psychological Science: A Journal of the
Association for Psychological Science,6, 3–5. doi:10.1177/
Buss, D. M., & Schmitt, D. P. (1993). Sexual strategies theory: An
evolutionary perspective on human mating. Psychological Review,
100, 204–232. doi:10.1037/0033-295X.100.2.204
Centraal Bureau voor de Statistiek. (2019). Honderd jaar alleenstaan-
den. Retrieved May 9, 2019, from Centraal Bureau voor de
Statistiek website:
Copen, C. E., Daniels, K., Vespa, J., & Mosher, W. D. (2012). First
marriages in the United States: Data from the 2006–2010 National
Survey of Family Growth. National Health Statistics Report,49,
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,
DePaulo, B. (2017). The urgent need for a singles studies discipline.
Signs: Journal of Women in Culture and Society,42, 1015–1019.
Green, P., & MacLeod, C. J. (2016). SIMR: An R package for power
analysis of generalized linear mixed models by simulation. Meth-
ods in Ecology and Evolution,7, 493–498. doi:10.1111/2041-
Hobbs, M., Owen, S., & Gerber, L. (2017). Liquid love? Dating apps,
sex, relationships and the digital transformation of intimacy. Jour-
nal of Sociology,53, 271–284. doi:10.1177/1440783316662718
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, 995–1006.
Kuznetsova, A., Brockhoff, P., & Christensen, R. (2017). lmerTest
package: Tests in linear mixed effects models. Journal of Statisti-
cal Software,82, 1–26. doi:10.18637/jss.v082.i13
Lancaster, K. (1990). The economics of product variety: A survey.
Marketing Science,9, 189–206. doi:10.1287/mksc.9.3.189
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,
169–182. doi:10.1109/TPC.2008.2000342
Lenton, A. P., & Stewart, A. (2008). Changing her ways: The number
of options and mate-standard strength impact mate choice strategy
and satisfaction. Judgment and Decision Making,3, 501–511.
Oberski, D. (2014). lavaan.survey: An R package for complex survey
analysis of structural equation models. Journal of Statistical Soft-
ware,57, 1–27.
Patall, E. A., Cooper, H., & Robinson, J. C. (2008). The effects of
choice on intrinsic motivation and related outcomes: A meta-
analysis of research findings. Psychological Bulletin,134,
270–300. doi:10.1037/0033-2909.134.2.270
R Core Team. (2013). R: A language and environment for statistical
computing. Vienna, Austria: R Foundation for Statistical Comput-
ing. Retrieved from
Robinson, H. (2018). Tinder: The marketplace for love. Critical
Reflections: A Student Journal on Contemporary Sociological
Issues,6. Retrieved from
Rosseel, Y. (2012). lavaan: An R package for structural equation mod-
eling. Journal of Statistical Software,2, 1–36.
Scheibehenne, B., Greifeneder, R., & Todd, P. M. (2010). Can there
ever be too many options? A meta-analytic review of choice over-
load. JournalofConsumerResearch,37, 409–425. doi:10.1086/
8Social Psychological and Personality Science XX(X)
Schwartz, B. (2004). The paradox of choice: Why more is less.
Retrieved from
Shephard, R. W., & Fa¨re, R. (1974). The law of diminishing returns.
Zeitschrift Fu
¨r Nationalo
¨konomie,34, 69–90. doi:10.1007/
Smith, C. (2018). 45 Interesting Tinder statistics. Digital Stat Articles.
Retrieved from
Wu, P.-L., & Chiou, W.-B. (2009). More options lead to more
searching and worse choices in finding partners for romantic rela-
tionships online: An experimental study. Cyberpsychology &
Behavior: The Impact of the Internet, Multimedia and Virtual
Reality on Behavior and Society,12, 315–318. doi:10.1089/cpb.
Zeileis, A., Leisch, F., Hornik, K., & Kleiber, C. (2002). An R Package
for testing for structural change in linear regression models. Jour-
nal of Statistical Software,7, 1–38.
Author Biographies
Tila M. Pronk is Assistant Professor at Tilburg University. Her work
focuses on romantic relationships. For example, she studies (online)
dating and forgiveness, as well as the impact of individual differences
like self-control on relationships.
Jaap J. A. Denissen is Professor at Tilburg University. His work
focuses on the interface between personality psychology, social psy-
chology, and developmental psychology. Broadly, he studies transac-
tions between persons and their environment.
Handling Editor: Vivian Zyas
Pronk and Denissen 9
... In this context, experimental studies using dating app paradigms have confirmed that when presented with a large set of dating app profiles, individuals reject more options, are less satisfied with their choices, and reverse them more than when exposed to smaller sets (D'Angelo & Toma, 2017;Pronk & Denissen, 2020;Wu & Chiou, 2009). Yet, we lack evidence on the effects of partner abundance on fear of being single. ...
... Experimental studies have shown that induced partner abundance versus scarcity has undesired consequences on relationships in that it increases infidelity intentions Arnocky et al., 2016) and partner demands (Locke et al., 2020). Furthermore, excessive choice on dating apps is detrimental to decision-making in the way that individuals reject more options, are less satisfied with their choices, reverse them more when presented with large sets of dating app profiles compared to smaller choice sets (D'Angelo & Toma, 2017;Pronk & Denissen, 2020;Wu & Chiou, 2009). ...
... Furthermore, dating app use had a direct effect on fear of being single. These results extend earlier work on choice overload on dating apps (e. g., Pronk & Denissen, 2020) by accounting for fear of being single as an outcome. Thus, the results confirm that not only partner scarcity but also partner abundance adversely affects the confidence to secure a relationship (Taylor, 2013). ...
Full-text available
Dating apps advertise with high availability of potential partners because users seem to prefer extensive choice. However, on the basis of consumer decision making research, we theorized that such excessive choice could have adverse effects, specifically on fear of being single, self-esteem, and partner choice overload. In Study 1, a survey with 667 adults between 18 and 67, dating app use was associated with an increased perception that the number of potential partners is numerous which, in turn, predicted higher fear of being single. In Study 2, we replicated the positive effect of partner availability on fear of being single in an experimental design with 248 adults between 18 and 38. We experimentally induced low, moderate, or high partner availability by assigning 11, 31, or 91 dating app profiles of allegedly available potential partners to participants. High (compared to low) partner availability increased fear of being single, decreased participants’ state self-esteem, and increased partner choice overload. Findings demonstrate pitfalls of excessive swiping on dating apps and extend choice overload literature by revealing effects on novel outcomes.
... Nasuprot navedenim podacima i istraživanjima, postoje i pretpostavke o tome da dostupnost sajtova za upoznavanje i olakšano pronalaženje partnera utiču na povećan broj samaca. Prema Pronku i Denisonu (Pronk, and Denissen, 2020), na osnovu empirijskog istraživanja koje su izvršili, konstantni pristup ljudi potencijalnim partnerima čini ljude više pesimistima i sklonim odbijanju. Nakon tri studije koje su izvršili kombinijući kvantitativni i kvalitativni pristup, zaključili su da do ovakvih okolnosti dolazi zbog pada zadovoljstva usled konstantog gledanja fotografija 3 i porasta pesimizma u vezi sa uspehom zabavljanja sa povećanjem vremena provedenog na sajtovima za upoznavanje. ...
... U vezi sa tim, Tinder nudi mogućnost plaćanja korisnima koji žele da njihova fotografija bude prikazana među prvima(Pronk, and Denissen, 2020). ...
Full-text available
The basic assumption of this paper is that the Internet has taken on a "matchmaking role" in creating emotional and marital relationships that are determinated with matching algorithms on the one hand, and virtual dating on the other. Therefore, our goal is to examine the basic social factors that contributed to the Internet's role in this regard. Accordingly, we will consider the potential consequences of such a transformation. First, we will analyze broader social processes that lead to the extraction of people from primary and direct relationships and to their re-rooting with the help of technological intermediaries. In this way, new communities based on the network principle are re-created, with the consequences of the building of friendly and emotional relationships. A revolutionary change of Internet technology is the possibility of making close connections with people who did not know each other before. In this regard, we will analyze the circumstances that led to the sphere of intimate relations becoming the most technologically mediated social sphere in the 21st century, because the Internet has become an irreplaceable, possible sovereign middleman of love affairs. The conclusion is that algorithms, not the "game of fate", will determine who will produce offspring with whom, in the future.
... It is possible that, contrary to our assumption, women who are active Bumble users may be casual daters who prioritize meeting many people over expressing in-depth self-disclosures. However, taking the findings of H4 and H5 together, it seems more likely that active Bumble users are interested in developing a relationship but may experience exhaustion related to high levels of romantic opportunities (Pronk & Denissen, 2020), leading them to be more selective about the interactions they invest in, while dedicating little self-disclosure to partners they are likely to reject. Meanwhile, those who spend less time on Bumble may be more willing to self-disclose, even with partners whom they ultimately reject. ...
Dating apps are an increasingly common element of modern dating, yet little research describes users’ experiences rejecting potential partners through these apps. This study examines how female Bumble users reject potential partners online in relation to self-disclosure, perceived partner disclosure, pre-rejection stress, and app usage. To investigate these issues, we conducted an online survey of 419 female Bumble users who had recently rejected someone through the app. Results revealed that women on Bumble employ ghosting strategies far more often than confrontational rejection and suggest that the degree to which women self-disclose, perceive a partner’s self-disclosure, and experience pre-rejection stress may impact their rejection strategies. This study informs the hyperpersonal model by demonstrating that reciprocal disclosure may characterize online dating interactions—even in relationships that fail to reach the face-to-face stage. However, results also broach the possibility of communication burnout in online dating, in which some users may lessen self-disclosure after extensive app usage.
... The effect of a decrease in motivation when too many options are available is called the choice-overload effect (Iyengar & Lepper, 2000) or the too-muchchoice effect (Scheibehenne et al., 2010). This effect is found in various domains like online dating (Pronk & Denissen, 2019) or consumer research (for an overview, see Scheibehenne et al., 2010). When comparing six choice options with a number of 24-30 choice options for rewards as a motivation to write an essay in class, researchers found that a smaller number was leading to higher motivation, satisfaction, and better essays (Iyengar & Lepper, 2000). ...
Full-text available
Based on the Self‐Determination Theory, the choice effect states that offering learners choice options independent from their relevance for learning will increase the perception of autonomy and intrinsic motivation as well as the learning performance. However, an increasing number of choice options can also have detrimental effects. This study examines the influence of the number of choice options on learners' performance. In the first experiment (N = 208) with a one‐factorial between‐subjects design with the factor number of choice options (two to six choice options) and an additional control group without a choice, the motivation, autonomy, cognitive load, and learning scores were measured. Results revealed that three to five choice options increased learning‐relevant variables most. In the second experiment (N = 180) with a one‐factorial between‐subjects design, two, four, or six choice options as experimentally relevant groups found in Experiment 1 were tested again to validate the findings and test mediations derived from the data of Experiment 1. Results reveal that the increase in learning from two to four options is mediated by an increase in decisional autonomy, whereby the decrease in learning from four to six choice options is mediated by a decrease in affective autonomy.
... A study focusing on women's sexual well-being determined that being in a committed relationship, having exclusive sex, having greater sexual agency, and having stronger desire all play a role in sexual well-being (Kaestle & Evans, 2017). However, online dating apps are known for leading to casual sex (Kallis, 2020;Sumter et al., 2017;Bonner-Thompson, 2020;van de Wiele & Tong, 2014), and choice overload on online dating apps was shown to increase rejection of potential partners (Pronk & Denissen, 2020). With online dating, it is easier than ever to meet new potential partners, and it was shown to facilitate the desire for long term commitment (Sharabi & Timmermans, 2020), but this doesn't guarantee sexual relationships to be satisfying. ...
Full-text available
The aim of this study was to investigate the connection of problematic online dating (POD), problematic social media (PSM), and problematic online sexual behaviors (POSB), with body esteem and sexuality. Previous research focused either on the impact of media on body esteem, or the impact of body esteem on sexuality. Yet, online media tends to be sexually self-objectifying, requiring more studies on their impact on body esteem and sexuality. In this study, a sample of 825 participants aged between 18 and 54 years old (M = 25, SD = 9.7), with 80% female participants, completed an online survey assessing POD, PSM, POSB, sex addiction, cognitive distractions during sexual intercourse, body esteem, sexual desire, sexual esteem, sexual depression, and sexual preoccupation. POD, PSM, and POSB were significantly correlated (r = 0.14, r = 0.35, r = 0.18). POD was linked to sexual depression (β = 0.10, p < 0.001), PSM was linked to body esteem (β = − 0.17, p < 0.001) and to cognitive distractions (β = 0.19, p < 0.001), and POSB was linked to sexual esteem (β = 0.14, p < 0.001). The mediation model indicated a significant indirect effect of body esteem and cognitive distractions between PSM and sexual depression (β = 0.15, Z = 4.39, p < 0.001), and for PSM and sexual esteem (β = − 0.13, Z = − 3.78, p < 0.001). This study highlighted the importance of studying outcomes of POD, PSM, and POSB on real-life sexual experiences, and to what extent body esteem and cognitive distractions were implicated. Further research is necessary on the impact of POD on sexual well-being and use of online sexual activities in diverse SM platforms.
... New media technologies offer access to more potential dates, permit encounters with people who we would not normally meet in our day-to-day lives, allow the use of computer-mediated communication to learn a wide range of facts about partners before meeting them in person, increase the ease with which affection or sexual preferences can be expressed, and offer diverse tools for negotiating stages of their love/sex relationships (Finkel, Eastwick, Karney, Reis & Sprecher, 2012;Meenagh, 2015). However, new media technologies also have downsides, such as the gradual feeling of discontent and pessimism about finding a mate (Pronk & Denissen, 2019), gamification of relationships, lack of romance and empathy on dating apps, and a growing use of behaviors like "ghosting", "slow fading", "benching", "breadcrumbing" or "haunting" (Cook, 2020). These behaviors illustrate how people are using technologies to flirt, initiate, maintain or end relationships. ...
Full-text available
The present study is part of a large study analyzing the prevalence of ghosting and breadcrumbing in sample of Spanish adults aged between 18 and 40 years. The study was split into different papers to better organize and understand the data obtained. The present paper investigated the prevalence of ghosting and breadcrumbing and associations between ghosting and breadcrumbing behavior and online dating practices. The results showed that half the participants were unfamiliar with the terms ghosting and breadcrumbing. However, approximately two in every 10 participants reported having experienced and initiated ghosting, and slightly more than three in every 10 participants had experienced or initiated breadcrumbing in the last 12 months. Regression analyses showed that the use of online dating sites/apps, more short-term relationships, and practicing online surveillance increase the likelihood of experiencing, as well as initiating, ghosting and breadcrumbing.
The personalization of digital environments is becoming ubiquitous due to the rise of AI-based algorithms and recommender systems. Arguably, this technological development has far-reaching consequences for individuals and societies alike. In this article, we propose a psychological model of the effects of personalization in digital environments, which connects personalization with motivational tendencies, psychological needs, and well-being. Based on the model, we review studies from three areas of application—news feeds and websites, music streaming, and online dating—to explain both the positive and the negative effects of personalization on individuals. We conclude that personalization can lead to desirable outcomes such as reducing choice overload. However, personalized digital environments without transparency and without the option for users to play an active role in the personalization process potentially pose a danger to human well-being. Design recommendations as well as avenues for future research that follow from these conclusions are being discussed.
This article reviews evidence for the social compensation hypothesis of online dating, according to which individuals who experience challenges with traditional dating gravitate towards and benefit from online dating. Three categories of psychosocial vulnerabilities that interfere with the initiation of romantic relationships are identified: 1) internalizing symptoms (i.e., anxiety, depression); 2) rejection sensitivity; and 3) attachment insecurity (i.e., anxiety, avoidance). The literature shows positive associations between anxiety, depression, rejection sensitivity, and attachment anxiety (but not avoidance) and online dating use. But significant lacunae exist in understanding the relational and wellbeing outcomes experienced by individuals with psychosocial vulnerabilities, or of the mechanism through which these vulnerabilities cause enhanced use of online dating. A detailed agenda for future research is proposed.
People's choices for specific romantic partners can have far reaching consequences, but very little is known about the process of partner selection. In the current study, we tested whether a measure of physiological arousal, pupillometry (i.e., changes in pupil size), can predict partner choices in an online dating setting. A total of 239 heterosexual participants took part in an online dating task in which they accepted or rejected hypothetical potential partners, while pupil size response was registered using an eye tracker. In line with our main hypothesis, the results indicated a positive association between pupil size and partner acceptance. This association was not found to depend on relationship status, relationship quality, gender, or sociosexual orientation. These findings show that the body (i.e., the pupils) provides an automatic cue of whether a potential partner will be selected as a mate, or rejected.
We propose a boundedly rational model of choice where agents eliminate dominated alternatives using a transitive rationale before making a choice using a complete rationale. This model is related to the seminal two-stage model of Manzini and Mariotti (2007), the Rational Shortlist Method (RSM). We analyze the model through reversals in choice and provide its behavioral characterization. The procedure satisfies a weaker version of the Weak Axiom of Revealed Preference (WARP) allowing for at most two reversals in choice in terms of set inclusion for any pair of alternatives. We show that the underlying rationales can be identified from the observable reversals in the choice. We also characterize a variant of this model in which both the rationales are transitive
Full-text available
One of the frequent questions by users of the mixed model function lmer of the lme4 package has been: How can I get p values for the F and t tests for objects returned by lmer? The lmerTest package extends the 'lmerMod' class of the lme4 package, by overloading the anova and summary functions by providing p values for tests for fixed effects. We have implemented the Satterthwaite's method for approximating degrees of freedom for the t and F tests. We have also implemented the construction of Type I - III ANOVA tables. Furthermore, one may also obtain the summary as well as the anova table using the Kenward-Roger approximation for denominator degrees of freedom (based on the KRmodcomp function from the pbkrtest package). Some other convenient mixed model analysis tools such as a step method, that performs backward elimination of nonsignificant effects - both random and fixed, calculation of population means and multiple comparison tests together with plot facilities are provided by the package as well.
Full-text available
In Liquid Love Zygmunt Bauman argued that the solidity and security once provided by lifelong partnerships has been 'liquefied' by rampant individualisation and technological change. He believes internet dating is symptomatic of social and technological change that transforms modern courtship into a type of commodified game. This article explores the experiences of users of digital dating and hookup applications (or 'apps') in order to assess the extent to which a digital transformation of intimacy might be under way. It examines the different affordances provided by dating apps, and whether users feel the technology has influenced their sexual practices and views on long-term relationships, monogamy and other romantic ideals. This study shows that dating apps are intermediaries through which individuals engage in strategic performances in pursuit of love, sex and intimacy. Ultimately, this article contends that some accounts of dating apps and modern romantic practices are too pessimistic, and downplay the positives of 'networked intimacy'.
Full-text available
This paper introduces ideas and methods for testing for structural change in linear regression models and presents how these have been realized in an R package called strucchange. It features tests from the generalized fluctuation test framework as well as from the F test (Chow test) framework. Extending standard significance tests it contains methods to fit, plot and test empirical fluctuation processes (like CUSUM, MOSUM and estimatesbased processes) on the one hand and to compute, plot and test sequences of F statistics with the supF, aveF and expF test on the other. Thus, it makes powerful tools available to display information about structural changes in regression relationships and to assess their significance. Furthermore it is described how incoming data can be monitored online.
Full-text available
This paper introduces the R package lavaan.survey, a user-friendly interface to design-based complex survey analysis of structural equation models (SEMs). By leveraging existing code in the lavaan and survey packages, the lavaan.survey package allows for SEM analyses of stratified, clustered, and weighted data, as well as multiply imputed complex survey data. lavaan.survey provides several features such as SEMs with replicate weights, a variety of resampling techniques for complex samples, and finite population corrections, features that should prove useful for SEM practitioners faced with the common situation of a sample that is not iid.
Full-text available
Maximum likelihood or restricted maximum likelihood (REML) estimates of the parameters in linear mixed-effects models can be determined using the lmer function in the lme4 package for R. As for most model-fitting functions in R, the model is described in an lmer call by a formula, in this case including both fixed- and random-effects terms. The formula and data together determine a numerical representation of the model from which the profiled deviance or the profiled REML criterion can be evaluated as a function of some of the model parameters. The appropriate criterion is optimized, using one of the constrained optimization functions in R, to provide the parameter estimates. We describe the structure of the model, the steps in evaluating the profiled deviance or REML criterion, and the structure of classes or types that represents such a model. Sufficient detail is included to allow specialization of these structures by users who wish to write functions to fit specialized linear mixed models, such as models incorporating pedigrees or smoothing splines, that are not easily expressible in the formula language used by lmer.
Online dating is often lauded for improving the dating experience by giving singles large pools of potential partners from whom to choose. This experiment investigates how the number of choices online daters are given, and whether these choices are reversible, affects romantic outcomes. Drawing on the choice overload and decision reversibility theoretical frameworks, we show that, a week after making their selection, online daters who chose from a large set of potential partners (i.e., 24) were less satisfied with their choice than those who selected from a small set (i.e., 6), and were more likely to change their selection. While choice reversibility did not affect daters’ satisfaction, those who selected from a large pool and had the ability to reverse their choice were the least satisfied with their selected partner after one week. The results advance understanding of how media features related to choice affect interpersonal evaluations.
The r package simr allows users to calculate power for generalized linear mixed models from the lme4 package. The power calculations are based on Monte Carlo simulations. It includes tools for (i) running a power analysis for a given model and design; and (ii) calculating power curves to assess trade-offs between power and sample size. This paper presents a tutorial using a simple example of count data with mixed effects (with structure representative of environmental monitoring data) to guide the user along a gentle learning curve, adding only a few commands or options at a time.