Empathizing personality traits for online dating applications
SUBMITTED IN PARTIAL FULLFILLMENT FOR THE DEGREE OF MASTER OF SCIENCE
ROB KOPPENAAL
STUDENT NUMBER: 11663170
M
ASTER
I
NFORMATION
S
TUDIES
H
UMAN-
C
ENTERED
M
ULTIMEDIA
F
ACULTY OF
S
CIENCE
U
NIVERSITY OF
A
MSTERDAM
10 October 2018
1st Supervisor 2nd Supervisor
Dr. Frank Nack Dr. Daniel Buzzo
Empathizing personality traits for online dating applications
Could personal questions increase traits of an individual in an online dating application?
Rob KoppenaalUniversity of Amsterdam Amsterdam, Netherlands
rob.koppenaal@student.uva.nl
ABSTRACT
Online dating apps are limited by lack of evaluating personality character traits [26]. To face this problem and increase success rates, dating apps could have options for statements/questions that are more personal during
conversation, rather than restrict the acquaintance to small talk [26]. This research focused on the question: Could personal questions increase traits of an individual in an online dating application?
In a controlled chatting conversation, 22 participants asked factual questions to each other, while another 22 participants asked personal questions. After the conversations, all 44 participants filled in two questionnaires: one as they saw their own personality and one as they felt for their chatting partner.
It was expected that increased knowledge on personality would result in closer resemblance on estimation of own and the other participants personality, but also on more explicit judgement (more honesty). Analyzing the data revealed no difference between both groups. Perhaps some questions would be significant if the sample size would have been bigger. Besides that, ten minutes of online chatting is too short to judge someone’s personality.
KEYWORDS
Online Dating, Personality, Connection
1 INTRODUCTION
Internet has facilitated the creation of social media like Facebook1, Instagram2 and WhatsApp3. In line with this trend
towards social media, online dating sites are popular as well. Although online dating has much to offer, there are still opportunities for further enhancements. One of the challenges of online dating is mentioned in several papers: it is tough to evaluate personality traits in the context of online dating: there is a lack of social cues that only exist offline [22, 23, 26].
When two people meet in a bar they have many social cues from the first second: body language, tone of voice, facial expressions, smell, touch, and a clear representation of
someone’s looks, instead of a few pictures that may be outdated.
1https://www.facebook.com/ 2https://www.instagram.com/ 3https://www.whatsapp.com/
Due to the lack of these non-verbal signals in online dating, people can evaluate one’s personality less accurately [11]. Instead all the first impression in online dating applications are gained through pictures, a small biography and messages. Besides that, the contact between the two online daters is often asynchronous (people do not send messages immediately but there is a delayed response) instead of synchronous
communication in the bar or the club. This results in people that can think about the message that one wants to send, what can give a different impression than meeting offline [11].
There is also a difference in mindset when people spend time at online dating sites. According to Hamilton, online dating sites are designed to encourage shopping-like strategies [11]. People are encouraged to look for people in a certain age group or height instead of directing to personality. However, these features have little impact for the quality of a relationship in terms of the duration of a relationship. Nevertheless, by looking for the perfect product in an environment where there are many choices, people have the tendency to make bad decisions. It is hard for people to distinguish relevant characteristics that are important for them from those that they would categorize as irrelevant.
People spend time on online dating sites for several reasons. One is to meet a partner for a potential serious relationship. A second one is for casual sex. Nevertheless, for both types of relationships, people have the desire to evaluate the others personality accurately [26].
A dating site that helps users to evaluate someone’s personality better is desired. In scientific research there is a knowledge gap for a solution to help online daters evaluate personality character traits [26]. According to Zytko et all., a clear interface that enables transfer of personality characteristics by text, could be a promising solution. Users could talk about specific topics and thus create a deeper personal connection with each other [9, 13]. A possible solution to counter this problem are statements/questions that get personal, rather than small talk [1]. However, an experiment as this is never conducted in context of online dating. In this thesis an approach is described that addresses this problem. The aim is to answer the following research question: Could personal questions increase traits of an individual in an online dating application?
The remainder of this thesis is structured as such. In the related work section, an overview is provided on already established work on supporting the better understanding of personality traits in dating sites. The method section elaborates
on the set-up of the experiment and the personality questionnaire that is used to describe the traits of other participants as well as themselves. The discoveries of the experiment are shown in the result section. Finally, there is a discussion and conclusion that includes a redesign proposal for dating sites based on the findings of this experiment.
2 RELATED WORK
2.1 Evaluating personality and selection
process
A requirement to know personality characteristics is to understand various methods people use to evaluate each other.
Personality traits: A well-used model of personality is the
five factor model, or the Big Five personality traits [5]. The five elements are known by the acronym OCEAN [2].
• Openness to experience: How curious and adventurous is someone?
• Conscientiousness: It is the balance between organizing and self-discipline in contrast to flexibility and easygoing. • Extraversion: Is one more extrovert or introvert? • Agreeableness: One can be more cooperative and willing
to come to a consensus, while on the other hand one can be more suspicious and can be a challenge to convince. • Neuroticism: One can be more sensitive for undesirable
emotions such as anger, anxiety and depression. Or one can be more emotionally stable.
Most users approach attractiveness with a cost-benefit analysis: physical attractiveness, demographic, lifestyle and personality traits, and the type of relationship one ones [22, 23, 26]. These parameters are vital for both a long-term relationship, as well as casual sex [26]. Zytko et al., distinguishes four other components of attractiveness.
Physical attractiveness: Personality is not the only
property that people look at. It matters a lot in online dating how handsome someone is. For men this seems more important than for women. This can be explained by evolutionary theory [4]. Males are turned on by females who are physical attractive, like a symmetrical face: unconsciously it is a sign of good genes and fertility. On the other hand, females find it important that a male has social status, confidence, and can protect her and her baby. Therefore, looks matter less for females [4].
Demographic traits: Although some people claim that
ethnicity does not matter for them, their behavior shows differently [7]. Black people are more likely to find black people attractive, similarly white people prefer Caucasians. However, demographic traits are more than the pigment of our skin. It is also about height, weight, age, religion, salary, and so on. People find it attractive to have a partner who has similar demographic traits [8]. Science regarding to online dating, shows that people are more willing to send messages to each other if one has similar demographic characteristics [7, 12].
Lifestyle traits: Most people look for a partner who’s
lifestyle is similar to them [26]. Attributes are: attitude, values and interests. E.g. ‘Does the other person smoke? Can I still eat
meat when the other person is a vegetarian? Does he or she enjoys adventure holidays?’
Relationship goal: What is the intention of the online
dater? Is it for a romantic long-term relationship or a short-term sexual experience [4]?
Although there are operationalized definitions of personality it is more subjective than in comparison to objective attributes like height, religion or if one smokes [5]. In online dating applications there is a focus on those objective attributes. However, when people are asked what they want to know about someone, they want to know if one has humor or if someone is social. The exclusive use of profile pictures and messaging makes it tough for online daters [26].
2.2 Increasing personality characteristics
In traditional dating sites people can present many details. For example: if they smoke or not, what their favorite food is, and how tall they are. However, with popular dating applications like Tinder4, the system does not ask thisinformation from their users, even though there are many people that use Tinder for a serious relationship [20]. With pictures and a short description of maximum 500 characteristics people present themselves. Nevertheless, according to Zytko et al., messaging has the most influence for people to determine if they want to go on a date with someone [16, 24].
In 2016, Zytko and coworkers performed an investigation among 34 online dating coaches [26]. These dating coaches do not only have their own online dating experience, but they also have references of their clients. Online dating coaches discuss the selection of pictures for the profile, the biography text and the conversations of their clients. By selecting representative profile pictures and a biography description, individuals can present themselves in an unrealistic positive way. The dating coaches concluded that meeting people face-to-face is always better than chatting online regarding evaluation [26].
According to Zytko et all., an interface that enables transfer of personality characteristics by text could form an appropriate solution. A deeper connection that is more personal can be created if online daters talk about specific activities [9, 13].These activities can be problem-solving interactions: interactions where users have contrasting attitudes and work towards consensus, or where both users have to cooperate to finish an assignment. The same applies to an opinion of a
statement/question that is more personal, rather than small talk [1]. These questions are found by the work of Aron [1] (also see Appendix B). A few examples are: ‘Would you like to be famous? In what way?’, ‘For what in your life do you feel most grateful?’ or ‘What is your most terrible memory?’. These are questions that lead to more closeness and intimacy, as shown by Aron in an experiment where 166 participants evaluated in a face to face session of 45 minutes where people went through those questions.
4https://tinder.com/
Another possible solution to increase traits of individuals is gaming [26]. In games, people participate in a problem-solving experience with a potential partner. Online games like World of Warcraft5 and Second Life6 have been ways where people have
found their marriage partner [10]. However, people do not intend to play a game like World of Warcraft to date someone. To play a game like World of Warcraft takes too much
investment to find someone. Perhaps a simple short game inside a dating application could be a good alternative.
In summary: questions that get more personal is perhaps a good method to increase personality characteristics. It is important to test this, because online dating and dating face-to-face has its differences: there are no non-verbal sign, people cannot hear or see each other. The experiment described in this paper is the first test in context of online dating.
3 METHOD
To be able to answer the research question: "Could personal questions increase traits of an individual in an online dating application?" a test design that compares the outcome of two dating groups is applied, of which one investigates the potential partner with factual questions, whereas the other group makes use of the set of personal questions as suggested by Aron [1] (see also Appendix B). The main hypothesis is: when people give their opinion to a personal question, they can evaluate traits of an individual with more certainty. There are a set of hypotheses that are guiding the process:
H1: Questions which imply that one knows the dater for a
longer time, will never be asked. (8, 11, 22, 25, 28 and 31 of the list of 36 questions)
H2: Less personal question will be asked more frequently. H3: When time increases people will ask questions that are
more personal.
H4: Participants that give their opinion to personal questions
are more likely to indicate whether the other participant described their personality quite accurately or very accurately in comparison to participants that ask factual questions.
H5: For every item of the personality questionnaire the answers
will be more certain for participants that give their opinion to personal questions in comparison to participants that ask factual questions.
H6: For every item of the personality questionnaire the margin
that both participants agreed will be smaller for participants that give their opinion to personal questions in comparison to participants that ask factual questions.
Figure 1 shows a visual overview of the two dating groups. Only the questions that are personal are used as a possible solution to increase traits of individuals, because due to the budget and time limit of this research it was more feasible than writing and applying a dating game. The participants in Group
5https://worldofwarcraft.com 6https://secondlife.com/
Factual had to ask factual questions to each other, the participants in Group Personal asked personal questions.
Figure 1: Group Factual represents the group that had to ask factual questions. Group Personal should ask personal questions.
3.1 Sample
As pilot experiment a convenience sample of 5 persons was used. With the population of 714490 university and higher professional education students of the Netherlands (including singles and relationships), a confidence level of 95% and a margin error of 15%, a sample size of 43 participants is necessary to make the sample size big enough [15]. This number should be doubled, because there are two groups in the experiment: Group Factual and Group Personal. Only 44 participants were recruited for this experiment. The participants were contacts of the researcher (i.e. convenience sampling). Because of limited time and budget there was no opportunity to recruit more and independent dating candidates. All participants were heterosexual: 22 males and 22 females.
3.2 Experiment
The instructions of the experiment are given in Appendix A. The four most important phases of the experiment are described in the underneath paragraphs.
3.2.1. Pairing of participants
The researcher approached the participants by convenience sampling. They were asked to participate in an online dating experiment. The researcher made sure that both participants did not know with whom they spoke. Furthermore, they could execute the experiment when they were at home, just like in online dating.
Participants who were paired got the instruction to ask personal questions or to ask factual questions to each other, depending on the group they were assigned to. They were told that they could elaborate on questions if they had the urge to do
so. Besides that, they were told that this experiment is in the context of online dating. They should pretend that they were chatting with a person they matched. They could chat for a maximum of 10 minutes or stop when they decided that there was no ‘chemistry’. When the time was over the researcher made this clear by sending the following message in the chatroom:
“---Researcher says---The conversation is ended. Please leave this chatroom and look at point 10 of the instructions. You have to fill in the personality questionnaire.”
3.2.2. Chatting
Participants chatted with each other with help of the tool
https://appear.in/. This is a website where one can set up a free video chat environment. However, when one turns off the microphone and cam functionality in the software, it is just an online chatroom. The researcher was the third person in the chatroom, who observed the experiment and could interrupt when necessary. The participants were aware of the observer. In an infield experiment an observer is not desired due to privacy concerns, however in this experiment there was an observer to get certainty that every user followed the provided questions and did not invent their own.
Participants got a list of 36 personal or 36 factual questions, depending on the group they were assigned to. The questions that were used in Group Personal are based on the 36 questions provided by the paper of psychologist Arthur Aron and his colleagues [1]. They are divided in three parts: varying from deep, to deeper and deepest. When the questions get deeper, and become more personal, the feeling of closeness increases.
The questions that were used in Group Factual are formed by the researcher itself, consisting of questions that people use in standard conversations, see Appendix C. The questions have in common that they are normally used in small talk conversations. The question are mostly closed-end questions that can be answered with a yes or no. They were tested by potential participants for their generalness and if people thought it were standard conversation questions.
Participants did not have the time to ask all the 36 questions. They had to choose from the list which one they would pick. Different questions were asked in every round. A ‘round’ stands for the timeframe in which a question is asked. Or in other words, when a participant starts the first question of the conversation it will be categorized as round 1. When the second question is asked, it will be categorized as round two etc. When a participant asked twice the same question, it was not included in the results.
3.2.3. Personality questionnaire
When the maximum time of 10 minutes was over, the researcher asked to the participants if they could fill in the questionnaire of Rammstedt & John, see Appendix D [17]. This is a short 10-item questionnaire that measures personality traits. Normally, this questionnaire is used to describe oneself, but for this experiment it is used to describe the other person.
Even though the questionnaire is a shorted version of the Big Five Personality model, according to Rammstedt & John it is significant for reliability and validity, even across multiple countries. A literature review of Kruyen et al., where they reviewed 2273 papers of personality traits concludes that the reliability of test like Rammstedt & John is only a little bit lower than the original and time consuming Big Five Personality model [14] and more accurate than similar simplified questionnaires [6].
Two adjustments were made to the questionnaire. Instead of ‘is generally trusting’ the researcher made it to ‘is generally trustworthy’. Besides that, the researcher deleted the word ‘few’ in the words ‘has few artistic interests’. The changes are based on the pretest, where these two statements created confusion among 2 of the 5 pre-test participants.
3.2.4. Check if personality matches
After the participants filled in the questionnaire to decide which personality characteristics the other user has. The results were shared with each other. In this part, just like during the whole experiment, the participants could not see each other. Then they had to decide how well the other participant
described them in general. Furthermore, they had to decide if the results that the other gave matches with their own personality.
After this step, the participants got individually interviewed by video chat or by text. Nobody got paid or reimbursed after the session. The researcher asked why they chose the specific questions of the list during their chatting experience?
4 RESULTS
The results section is separated in three sections. (1) An analysis of the interviews. (2) Patterns in the chatting behavior. (3) Results based on the personality questionnaire.
4.1 Results of the interviews
After the experiment the researcher held an interview with the participants and openly asked their reason for asking certain questions. Their arguments were described in own words but could be clustered. Therefore, the answers are clustered and presented in Table 1. This analysis shows the motivation that people have when they ask their specific questions. Perhaps interesting patterns can be discovered. As expected, participants of Group Personal were in general much more concerned about questions that would reveal the individual traits (16 people of Group Personal against 8 persons of Group Factual gave the reason that they want to know the opinion of certain questions to make an evaluation of someone’s personality).
Table 1: Analysis of the interviews that were conducted in Group Personal and Group Factual.
Reason to ask the question, because… Number of answers Group Personal Number of answers Group Factual
These questions are important for me and I want to know how the other person feels about it. Or the answer of these questions would probably give me a good feel about the others personality
16 8
I want to start off with a question that was not too personal and easy to break the ice
6 3
It was logical to hold the conversation 4 6
The question was fun 4 4
Random 1 2
Total arguments 31 23
4.2 Chatting behavior of participants
During the experiment, certain patterns were observed in the chatting experience. No hypothesis was made on the differences between male and female behavior in online dating. However, there were clear patterns and therefore the findings are included in the results: Table 2, 3 and Figure 2, 3.
Table 2 summarizes which gender started the conversation. The start of the conversation is in this experiment defined as the first person who asked a question of the list of 36 questions. Males were more likely to start the conversation in Group Personal, in Group Factual there was no clear difference.
Table 2: Gender difference for starting the conversation Group Personal versus group Factual
Gender Group Personal
Group Factual
Male 10 7
Female 1 4
H1 stated: Questions which imply that one knows the dater for
a longer time, will never been asked. (8, 11, 22, 25, 28 and 31 of the list of 36 questions). Figure 2 shows which questions are asked. All the questions of H1 are not asked, except for question 11.
H2: Less personal question will be asked more frequently.
Figure 2 displays the number of times each of the 36 questions were asked in Group Personal. One can see that more personal questions are asked less frequently. Table 3 is an overview of all the questions regarding the part to which they belong and the number of times they were asked. Even in Group Personal people tend to prefer less personal questions.
Table 3 shows a difference as well in how deep males and females ask questions. In Group Personal, males asked 22 questions and females 14 questions in the 11 conversations. Despite that females asked less questions, they asked deeper questions in general than males. Males ask in general more questions, and the questions that they ask are less deep than females.
Figure 2: The occurrence of the 36 questions in Group Personal and a visual overview to the part where every question belongs.
Table 3: Categorization of the three parts each consisting of 12 questions asked in Group Personal
Parts Total Number of asked questions Male Female Part 1: deep Question: 1-12 22 14 8 Part 2: deeper Question: 13-24 12 7 5 Part 3: deepest Question: 25-36 2 1 1
H3 stated: When time increases people will ask questions that
are more personal. Figure 3 shows a displacement from deep questions to deeper questions during the 4 rounds. There was a tendency of deeper questions during consecutive rounds.
0 1 2 3 4 5 6 7 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 A mou nt of qu es ti on s that are as ked The 36 questions
The 36 questions of Group Deep
Figure 3: The x-axis displays the rounds: the timeframes in which a question is asked. The y-axis shows the actual questions. As one can see, deeper questions are asked when the rounds increase.
4.3 Personality questionnaire
H4 stated: Participants that give their opinion to personal
questions are more likely to indicate whether the other participant described their personality quite accurately or very accurately in comparison to participants that ask factual questions. Figure 4 shows what participants thought of the correctness in personality how the other participant described them. Two participants (one in Group Factual and one in Group Personal did not gave an answer for this question. They were left out of the results. There was no significant difference in the accuracy of the other participant in both groups, because the answers in all groups are identical or differ only by one answer.
Figure 4: People could give their opinion on how well the other participant described them in a scale of very inaccurately to very accurately.
H5 stated: For every item of the personality questionnaire the
answers will be more certain for participants that give their opinion to personal questions in comparison to participants that ask factual questions. Figure 5 shows the average answer on every item of the personality questionnaire. The y-axis displays the measure of agreement: 2 for agree strongly and -2 for disagree strongly. The distinction between -2 till 2 is made, because it is easy to see how much the answers are separated from each other. In five occasions someone forgot to fill in an answer for one attribute in the personality questionnaire. These missing attributes are not used for the results. There was no significant difference between the average answers in both groups. A tailed distribution T-test was performed on a two-sample unequal variance (heteroscedastic), Table 4. Perhaps some questions, like question 3 and 5, would have a significant difference if the sample size would have been bigger.
Unfortunately, there were not any relations or pattern that could explain which questions contribute to the option that the other participant described the user correctly.
0 2 4 6 8 10 12 14
How do you think the other participant
described your personality in general?
Figure 5: For every attribute of the personality questionnaire, people answered from a scale of disagree strongly to agree strongly. The error bar is the standard deviation.
Table 4: T-test scores of Figure 5 for a two-tailed distribution T-test on a two-sample unequal variance (heteroscedastic). Question T-test 1 0,467083493 2 0,278869439 3 0,195350338 4 0,121357221 5 0,178709149 6 0,306940454 7 0,160730626 8 0,446026057 9 0,615107582 10 0,694630615
H6 stated: For every item of the personality questionnaire the
margin that both participants agreed will be smaller for participants that give their opinion to personal questions in comparison to participants that ask factual questions. Figure 6 displays the average margin for every item of the personality questionnaire that both participants agreed on. If they both agreed on the same item, the margin was 0. If one participant said that he strongly disagreed, and the other participant said that she strongly agreed, the margin was put to 4. Again, the 5 occasions that someone forgot to answer are not used for the results. There was no significant difference between the margins in both groups. A two-tailed distribution T-test was performed on a two-sample unequal variance (heteroscedastic), Table 5. Perhaps some questions, like question 5 and 10, would have a significant difference if the sample size would have been bigger.
Figure 6: The average margin for every attribute of the personality questionnaire. The error bar is the standard deviation.
Table 5: T-test scores of Figure 6 for a two-tailed distribution T-test on a two-sample unequal variance (heteroscedastic). Question T-test 1 0,901188127 2 0,10142889 3 0,723359168 4 0,438578628 5 0,055829942 6 0,357741815 7 0,836843398 8 0,558693715 9 0,536257483 10 0,176831308
5 DISCUSSION
The discussion is divided in four sections. (1) Challenging the null-hypothesis. (2) The interpretation of the results and how the results correspond to other scientific studies. (3) A critical evaluation of possible improvements for this experiment. (4) Suggestions for future research.
5.1 Challenging the null-hypothesis
The results from the chatting behavior indicate that participants in Group Personal were more interested in the personality of the other participant than the participants in Group Factual (Table 1). As argued before, the primary key to a successful dating app is to increase traits of individuals. To get to know each other, before physical dating.Men were more likely to start a conversation, Table 2. Online dating apps should take this in consideration. It is unclear whether this outcome will have influence on the further dating process. It could be that women often receive more messages then men and that woman are more in the habit that men start a conversation [19]. The difference in gender is less in Group Factual than in Group Personal. As personal questions have
-2 -1 0 1 2 1 2 3 4 5 6 7 8 9 10 Fro m dis ag ree st ro ng ly to ag ree st on gl y
The 10 questions of the personality questionnaire
Average answer to personality
questionnaire
Factual Deep 0 0,5 1 1,5 2 1 2 3 4 5 6 7 8 9 10 Th e av ara ge ma rg inThe 10 questions of the personality questionnaire
Margin
more empathize to traits of individuals and the fact that females consider personality more important than males [4]. This idea is confirmed with the finding that females asked more personal questions than males, Table 3.
In this experiment the interviews were interesting. Almost everybody said two things when the researcher asked what their opinion of the research was: chatting was fun, unfortunately that the time went so fast. The opinion of almost everybody says a lot about the experiment and is therefore extremely important. It is likely that 10 minutes of chatting is too short to evaluate someone’s personality.
Participants could chat for 10 minutes with each other before the researcher brought the conversation to an end. However, sometimes the researcher waited for one more minute if one participant was typing for a long time, but no answer had yet appeared.
Furthermore, in the research of Aron, people had to talk to each other for 45 minutes face-to-face. After that, they had to look each other in the eyes for 4 minutes, without saying anything [1]. This is different from online dating for various reasons. There was more time to ask questions. Besides that, people could see each other. All these elements are likely to make a better evaluation of someone. This is a good explanation of the differences in results between the experiment of Aron and the one that is described in this paper where participants could solely communicate by chat for the short time of 10 minutes.
However, it could also be that the personal questions of Aron had a distinction between positive and negative questions [1]. Obviously, most people do not ask negative questions at the initial contact of a conversation. Therefore, the researcher asked 6 participants if they could categorize the 36 questions of Group Personal in positive/neutral or negative questions. It turned out that the questions of Table 6 are the most negative questions of the 36 personal questions. As one can see there are many matches that correspond with Figure 2. Only questions 33 and 35 are categorized as negative and are never asked. Therefore, it is unlikely that this experiment was a research of positive and negative questions instead of factual and personal questions.
Table 6: Negative questions of the 36 questions of Group Personal (Only scores of 3 and higher), 6 participants.
Question People scores it negative
3 3 7 5 18 6 30 5 32 4 33 4 35 6
5.2 Interpretation results on the hypotheses
H1: Questions which imply that one knows the dater for a
longer time, will never be asked. Although participants could ask all questions from the list, it was clear that they avoided questions which imply that one knows the dater for a longer
time (Figure 2). Confronting potential partners with these questions in a first chat, could make the conversation
unnecessary complicated. Dating apps should consider this and avoid confronting questions that imply one knows the other person very well.
There was one exception, question 11: ‘Take four minutes antand tell your partner your life story in as much detail as possible.’ This question was asked only one time. The answer of the participant was a fictional humorous summary that took approximately 2 minutes to write. The researcher thought this question would never be asked nature because it is too much investment in context of online dating. In online dating people only get to know each other in a short timeframe, especially in a timeframe of 10 minutes for this experiment. However, one participant asked this question and explained that she chose it because she thought that if someone described his life story in four minutes, it gives a good impression of someone’s life.
H2: Less personal question will be asked more frequently. As
expected, questions in part 1 where more asked then the deeper questions of part 2 and 3 (Table 3). For dating apps, this is a serious topic to deal with. An interaction system must be human-centered designed, and therefore it is necessary to understand the user. A smart interface is designed in a way that encourages people to start with less personal questions instead of going straight to a really personal one. As explained later, more time could solve this problem.
H3: When time increases people will ask questions that are
more personal. To investigate whether people would loosen up when they spend more chatting time together, the experiments consisted of 4 possible rounds of questioning in each
conversation. As expected, when the rounds increased, the questions got deeper and more personal (Figure 3). Or in other words, if the time increases, questions of part 3 are more frequently asked. In fact, this is a reflection or normal life. When one gets a new college, he will explain what to do. In time one will tell more about it and elaborate on the subject. Dating apps should be aware of this normal life reflection and draw the user’s attention to this topic instead of going immediately to personal topics. People in Group Personal need time to feel comfortable with asking deeper questions. A possible implementation of an interface would be that the user gets aware that only after a certain amount of questions/time one can ask deeper questions that aim for a possibly better evaluation of personality.
H4: Participants that give their opinion to personal questions
are more likely to say that the other participant described their personality quite accurately or very accurately in comparison to participants that ask factual questions. Plotting all average answers of every item of the personality questionnaire, there is no significant difference (Figure 4). For dating apps this outcome is unreliable. Perhaps it is hopeful or just threatening. Hopeful, if it means that people are always honest when chatting with a potential partner, threatening as honesty is difficult to measure, and thus difficult to deal with.
H5: For every item of the personality questionnaire the answers
will be more certain for participants that give their opinion to personal questions in comparison to participants that ask factual
questions. It was expected that people in Group Personal had a stronger opinion because they are more certain about the personality of the other, instead of someone in Group Factual who has not a good picture of the others personality. In this limited experiment this difference did not show up (Figure 5). None of the ten items of the personality questionnaire turned out to be significant of the Two-tailed distribution T-test that was performed on two-sample unequal variance
(heteroscedastic). Perhaps some questions would have a
significant difference if the sample size would be bigger. Like H4, it is difficult to find honesty rules based on questionnaires. As honesty is the fundament of each relationship other strategies should be investigated.
H6: For every item of the personality questionnaire the margin
that both participants agreed will be smaller for participants that give their opinion to personal questions in comparison to participants that ask factual questions. According to the
hypothesis it is expected that people in Group Personal are more likely to have an idea of someone’s personality. Therefore, one would aspect that the average difference is low in Group Personal and high in Group Factual. However, when the results were plotted there was no difference (Figure 6). It is hard to interpret this outcome. Does this mean that even though people ask each other more personal questions they do not get to know each other? Or it might be that the time is too short to make an accurate evaluation of someone's personality. Future research should focus on an experiment where the chatting time has a longer duration.
5.3 Technical evaluation of the experiment
The evaluation of the experiment is classified in three categories. (1) points of improvement for the sample. (2) A transparent overview of points that could be improved before the participants had their conversation. (3) Issues during chatting. (4) Input of the participants retrieved from the interviews. (5) Technical improvements for future research.5.3.1 Sample
Participants that took part of this experiment were all friends of the Dutch researcher. This kind of ‘convenience sampling’ does not result in a realistic representation of the average user of online dating. As such, the range of age was between 19 and 32, almost all persons have a background at university or higher education and they had similar hobbies and interest as the researcher. Future researchers should try to contact dating sites and see if they could implement the design principles that will be stated later in this paper. Then one can test the experiment in a realistic setting. One should also keep in mind that (online) daters are a subpopulation of the Dutch society which also varies a lot. To find a representing group of participants, researchers should approach people at different events e.g. pop concerts, church gatherings, environment protest and so on. Another approach is the recruitment of different age groups.
Several techniques were conducted to exclude different variables. Everybody in this experiment was heterosexual thus the matching technique is used to make sure that 50% of the
participants was male and the other half was female. The distinction between heterosexuals and same-sex or people of the LGBT community is that they experience online dating
differently [11]. According to Smith & Duggan, women and gay men receive in general much more messages than men who are heterosexual [19]. The lack of messages can receive anxiety when men use a dating site [24]. It is interesting to see if these facts also have influence on the results when one performs the same experiment but with a variety of sexual preferences.
The second technique was based on random assignment of the participants. The researcher assigned participants in Group Factual or Group Personal. However, he did not put any bias in this process, because participants were assigned to the group with the minority of participants. In short: choices that were made by assigning were made on a logical basis of schedule and no other reason. Future researchers should try to perform randomized double-blind experiments. Of course, that could mean that two heterosexuals of the same gender are coupled, or homosexual persons to heterosexual persons. An interesting match, as the chatting persons would not know anything of the other.
5.3.2 Before the conversation
Three participants, all of Group Factual, were unable to use the chatroom tool https://appear.in/. The reason was that people had to conduct the experiment on their mobile phone because their computer was not working or unavailable. Therefore, the experiment was continued on WhatsApp7. Participants got the
instructions to make their profile picture invisible and delete their WhatsApp status, by doing so they could not have a first impression before the chatting would start. However, it could still have an influence on the participants, because chatting on smartphones has a different experience than chatting on a desktop. For example, usually people type faster with a keyboard than touchscreen [18].
The chat tool https://appear.in/ turned out to be not the best tool one could find on the market. There were some
complications, for example: one participant entered the chatroom with her webcam on for the first 10 seconds. Luckily, the other participant was not in the chatroom yet, otherwise their session had to be canceled. It would be better if there would be a tool were this would be impossible. It's unknown if the different question types result in different results when participants would use the camera functionality. Due to the increasement of communication channels it would be likely that this will help to evaluate personality traits [3].
Another point of improvement is that the chatroom environment should work in every browser and on any laptop. Unfortunately, there was one moment were https://appear.in/
did not work and people had to download another browser on the spot before they could start chatting.
It happened four times in Group Factual and three times in Group Personal that a participant had to wait in the chatroom before the other participant appeared for a period of 5 minutes
or longer. This due to slow reading, computer problems, or other issues. This can possible give the other participant a negative first impression ‘Why does it take so long?’. Perhaps participants could also feel stress, because they know that someone else is waiting in the chatroom and might lose patients. To start the experiment in a stressful condition can have influence on the way people chat. However, the researcher did not notice any stress that could be noticed in the chatting behavior.
Despite of most people who said that the instructions were crystal clear, the researcher got a few messages of participants that it was not clear for them if they could exclusively ask question or could give answers on the questions as well.
5.3.3 While Chatting
In Group Personal it happened twice that the researcher interfered the experiment to make sure that the participants stayed close to the questions. The reasons behind the interruption was that the conversation did not went personal anymore, because the participants asked informational facts to each other rather than personal questions. The group that had to ask personal questions should not ask too many factual
questions, otherwise the data could not be taken into the results. However, by disturbing the flow of the conversation, people are maybe distracted and do not make certain actions what could lead to different results [21]. The examples of the conversations where the diversion happened are listed below. They are both translations of two Dutch conversations.
Conversation 1:
Man: What roles do love and affection play in your life? Woman: let me thing, Ehm I would say an important role. I
find it important to share and receive love. Without I would definitely miss something in my life.
Man: why
Woman: it sounds a bit kitsch I know :D
Man: Haha yes sorry I am picky (In this sentence ‘Picky’ is
not a correct translation because in the Dutch he made a joke by writing the word ‘kitschkeurig’)
Woman: daaaamn haha would you not miss anything if there
was no love?
Man: Now you are asking me one! Haha. Yes of course, you are
abandoned as a child if you would answer no Anyway you are dating at the moment?
Interruption of the researcher Conversation 2:
Woman: What do you value most in a friendship?
Man: I think harmony is important in friendship, fun, playful
teasing and especially honesty. In this way you can always rely on each other and have a good time.
Woman: Sounds good! And did you find it in a friend/female
friend (or more)?
Man: Well, I am super boring person who is only busy with his
pc and big toe :p
Woman: Shit, now we go somewhere..
ANYWAY
So, your answer for the question is my pc’?
Our house, containing everything you own, catches fire. After saving your loved ones and pets, you have time to safely make a final dash to save any one item. What would it be? Why?
Interruption of the researcher
The diversion of conversation 1 happened when the man started to ask a logistical question about the relationship status of the woman. What is a pure factual question and not a personal one. The diversion of conversation 2 happened when the man gave an answer that was more focused on fun. The woman replied with a scenario that might could lead to a more personal connection, however that is unknown while the provided questions had a bigger chance to create a deeper connection.
Group Factual of this experiment had to ask factual questions to each other. They did, however in most conversations there was also humor involved, in general more than in Group Personal which was more on the serious/connection side of the spectrum. Perhaps it could be that in this experiment there was no comparison between factual and personal questions, but rather conversations with humor and conversations that are more personal. This possible validity gap must be considered by stating the conclusion.
Of the 22 conversations, 13 times it happened that the participants continued with chatting after the lime limit of 10 minutes was over. This happened five times in Group Factual and seven times in Group Personal. Normally they thanked each other for the nice time or said that they did not noticed how fast the time went. Nevertheless, one time it happened that the participants just continued their conversation. The researcher had to kindly remind them to stop. This doesn't necessarily mean that participants of personal conversations where more interest in the other. It is also hard to say if people thanked each other out of politeness or real empathic. However, according to the interviews every person liked to chat to the other participant what makes the option of real empathic more likely.
For two participants, both of group A, it took between 20 and 30 minutes to fill in and send the personality questionnaire to the researcher after chatting. When the researcher asked in the interviews why they were so slow, they answered that they thought there was no rush in the process. The participants said also that they still remembered which and why they chose the questions out of the list of 36 questions. Therefore, the possible impact of this, is probably not so big.
Two times somebody mentioned that there were difficulties due to the language barrier. Dutch was their native language and they had a few questions about the instructions that were in English. Two participants asked for help, nevertheless there could be more people that had difficulties but were too ashamed or shy to ask.
5.3.4 Interviews
In the interview that was conducted a few participants mentioned that they hesitated how much they could stay close to the questions or if they could elaborate on a topic. Although the
answer to this question was mentioned in the instructions, it was still unclear for some people.
During this experiment, the participants had to fill in a personality questionnaire to evaluate the other as well as themselves. It is possible that participants based their answers on the way they think they came across by the other participant. Afterwards, it was also a possibility that participants had to fill in what they considered their personality a few days before they took part of the experiment.
5.3.5 Technical improvements for future research
This research is conducted in a laboratory setting. Thus, conclusions of this research do not have to be exactly true for infield testing. For example, it is likely that people act differently when they know that nobody is watching their conversation instead of knowing that every message gets recorded. Another example is someone’s relationship status.
Furthermore, the laboratory setting could be improved if participants had the ability to sit in two separate rooms were the whole experiment was programmed out. This would have possibly prevented that a few people don’t have to wait for each other, and that the environment is the same.
Besides, it is interesting to see if it makes any difference when participants fill in the personality questionnaire a few days before the experiment starts. It could be that people are
influenced by the impressions that they get from the other participants. Nevertheless, it would be interesting to see how accurate people are when they evaluate their own personality and that of others by the 10-item personality questionnaire.
In this experiment it didn’t matter when people were in a relationship or not. This could have an influence on the results. Although, Zytko argues that even for casual sex, people want to evaluate someone’s personality, maybe they only want to know specific attributes [26]. Perhaps, people that have the intention of casual sex, only want to know how sexual free, easy or extrovert someone is. Future researches should handle these topics, because this experiment could not find specific patterns of questions that lead to these personality characteristics. If patterns are found a dating application could provide users a minimum of questions that lead to the quickest evaluation of someone.
6 CONCLUSION
The main research question during this experiment was: "Could personal questions increase traits of an individual in an online dating application?" The conclusion of this paper is that personal questions lead to the same evaluation of character traits as factual questions. Therefore, the hypothesis: ‘when people give their opinion to a personal question, they can evaluate traits of an individual with more certainty’ is rejected. Perhaps some questions would be significant if the sample size would have been bigger. Besides that, ten minutes of online chatting is too short to judge someone’s personality. When participants could chat longer, chances are that it would lead to different results. However, this should be tested in a new experiment.
6.1 Design Implementations
There are dating sites where one can answer questions that the dating site has provided, for example OKCupid8. However,
they use the answers of the questions in their algorithm to find a better match for the user, whereas this this experiment used questions to make a connection with an existing match. These questions should be provided in a specific order.
Out of the interviews is concluded that people find it comfortable to start a conversation with an ice-breaker. Both men and women do not want to start a conversation with a really personal question. However, after the first question people feel more comfortable to go deeper in the conversation by asking deeper questions to each other. Despite that females asked less questions, they asked deeper questions in general than males. Males ask in general more questions, and the questions that they ask are less personal than females. For a dating site this would mean that they should direct users to ask ice-breaker openings at the beginning of the conversation. Over a short time, the system could provide deeper questions, whereas the system could provide females earlier with more personal questions than men.
During this experiment men were more likely to start the conversation. This is logical because women tend to wait for an approach because they often receive more messages than men [22]. To prevent that women get overwhelmed with messages, it could be a solution that only women can start the conversation. An existing dating app that has successfully implemented this method is Bumble9.
A dating site should not provide the user questions that imply that one knows the dater for a longer time. Confronting potential partners with these questions in a first chat could make the conversation unnecessary complex.
Obviously, there are more ways to increase personality traits in online dating. Future researchers could partly overcome these shortcomings by using audio messages or video chats. They can show personality with more certainty, because it gives more communicational signals [3, 26]. Furthermore, Zytko stated that problem-solving interactions could express personality better [26]. This could be implemented in a game where both
participants have a problem and they can only solve the problem by working together. Or scenario’s that are provided to both online daters where they must give their opinion about.
Ten minutes was too short to evaluate someone's personality online. However, it can be interesting to see if a similar
experiment is done in the context of a speed dating event. Would personal questions increase personality characteristics? It seems likely that the best way to evaluate personality, is offline in a face-to-face setting [25, 26].
8https://www.okcupid.com/
REFERENCES
[1] Aron, A. et al. 1997. The experimental generation of interpersonal closeness: A procedure and some preliminary findings. Personality and Social Psychology Bulletin. 23, 4 (1997), 363–377.
[2] Big Five personality traits: https://en.wikipedia.org/wiki/Big_Five_personality_traits. Accessed: 2018-06-19.
[3] Brave, S. and Nass, C. 2009. Emotion in Human– Computer Interaction. January 2002 (2009), 53–68. DOI:https://doi.org/10.1201/b10368-6.
[4] Buss, D.M. and Schmitt, D.P. 1993. Sexual strategies theory: an evolutionary perspective on human mating. Psychological review. 100, 2 (1993), 204.
[5] Costa, P.T. and MacCrae, R.R. 1992. Revised NEO personality inventory (NEO PI-R) and NEO five-factor inventory (NEO-FFI): Professional manual. Psychological Assessment Resources, Incorporated.
[6] Credé, M. et al. 2012. An evaluation of the consequences of using short measures of the Big Five personality traits. Journal of Personality and Social Psychology. 102, 4 (2012), 874–888. DOI:https://doi.org/10.1037/a0027403. [7] Fiore, A.T. and Donath, J.S. 2005. Homophily in online
dating: when do you like someone like yourself? CHI’05 Extended Abstracts on Human Factors in Computing Systems (2005), 1371–1374.
[8] Fisher, H. 2016. Anatomy of Love: A Natural History of Mating, Marriage, and Why We Stray (Completely Revised and Updated with a New Introduction). WW Norton & Company.
[9] Gonzaga, G.C. et al. 2007. Similarity, Convergence, and Relationship Satisfaction in Dating and Married Couples. Journal of Personality and Social Psychology. 93, 1 (2007), 34–48. DOI:https://doi.org/10.1037/0022-3514.93.1.34.
[10] Hall, J.A. 2014. First comes social networking, then comes marriage? Characteristics of Americans married 2005–2012 who met through social networking sites. Cyberpsychology, Behavior, and Social Networking. 17, 5 (2014), 322–326.
[11] Hamilton, N.F. 2016. Romantic Relationships and Online Dating. Applied Cyberpsychology. Palgrave Macmillan UK. 144–160.
[12] Hitsch, G.J. et al. 2010. Matching and sorting in online dating. American Economic Review. 100, 1 (2010), 130– 163.
[13] Karney, B.R. and Bradbury, T.N. 1995. The longitudinal course of marital quality and stability: A review of theory, methods, and research. Psychological Bulletin. 118, 1 (1995), 3–34. DOI:https://doi.org/10.1037/0033-2909.118.1.3.
[14] Kruyen, P.M. et al. 2013. On the Shortcomings of Shortened Tests: A Literature Review. International Journal of Testing. 13, 3 (2013), 223–248. DOI:https://doi.org/10.1080/15305058.2012.703734. [15] Leerlingen, deelnemers en studenten; onderwijssoort,
woonregio:
http://statline.cbs.nl/StatWeb/publication/?PA=71450ned. Accessed: 2018-06-20.
[16] Ramirez, A. et al. 2014. When online dating partners meet offline: The effect of modality switching on relational communication between online daters. Journal of Computer-Mediated Communication. 20, 1 (2014), 99–114.
[17] Rammstedt, B. and John, O.P. 2007. Measuring personality in one minute or less: A 10-item short version of the Big Five Inventory in English and German. Journal of Research in Personality. 41, 1 (2007), 203–212. DOI:https://doi.org/10.1016/j.jrp.2006.02.001. [18] Sears, A. 1991. Improving touchscreen keyboards:
design issues and a comparison with other devices. Interacting with computers. 3, 3 (1991), 253–269.
[19] Smith, A. and Duggan, M. 2013. Online Dating & Relationships. Pew Research Center’s Internet & American Life Project. Online Dating & Relationships| Pew Research Center’s Internet & American Life Project.”, Pew Research Center Internet Project. (2013).
[20] Sumter, S.R. et al. 2017. Love me Tinder: Untangling emerging adults’ motivations for using the dating application Tinder. Telematics and Informatics. 34, 1
(2017), 67–78.
DOI:https://doi.org/10.1016/j.tele.2016.04.009.
[21] Wendel, S. 2013. Designing for behavior change: Applying psychology and behavioral economics. “ O’Reilly Media, Inc.”
[22] Zytko, D. et al. 2015. Enhancing Evaluation of Potential Dates Online Through Paired Collaborative Activities. the 18th ACM Conference. (2015), 1849–1859. DOI:https://doi.org/10.1145/2675133.2675184.
[23] Zytko, D. 2016. Enhancing Evaluation of Potential Romantic Partners Online. Proceedings of the 19th International Conference on Supporting Group Work - GROUP ’16. (2016), 517–520. DOI:https://doi.org/10.1145/2957276.2997030.
[24] Zytko, D. et al. 2014. Impression Management Struggles in Online Dating. Proceedings of the 18th International Conference on Supporting Group Work. (2014), 53–62. DOI:https://doi.org/10.1145/2660398.2660410.
[25] Zytko, D. 2016. Online Dating Coaches ’ User Evaluation Strategies. CHI Extended Abstracts on Human Factors in Computing Systems. (2016), 1337–1343. DOI:https://doi.org/10.1145/2851581.2892482.
[26] Zytko, D. et al. 2016. The Coaches Said...What?: Analysis of Online Dating Strategies Recommended by Dating Coaches. Proceedings of the 19th international conference on supporting group work. (2016), in press. DOI:https://doi.org/10.1145/2957276.2957287.
APPENDIX A: THE INSTRUCTIONS
Instructions (please read them carefully)
Thanks for joining this experiment and making your contribution to science. This chat will be recorded, but your answers will be used anonymously in a publication.
1. You are joining an online dating experiment. Imagine that you have a match on an online dating application. You will be chatting with your potential new date.
2. Nevertheless, you can only ask questions of the list underneath. Scan the 36 questions, however the goal is just to see them and not to remember them. While chatting you can keep this document open and look at the 36 questions again.
3. You can ask the questions randomly and don’t have to follow the chronological order. There is no rush in this experiment, you cannot finish all the 36 questions in time. Your conversation should be in English or Dutch.
4. If you have finished reading the 36 questions, you can go to the online chatroom. Please turn off the microphone and cam functionality at the bottom of the screen as quickly as possible, see picture. Also hold your finger in front of the camera for the first seconds, just to be sure.
5. You can keep chatting for a maximum of 10 minutes or when you decide there is no chemistry. When there is no chemistry say this in that chatroom: NO CHEMISTRY.
6. Be aware that the other participant maybe needs more time to read the questions and you should wait a bit.
7. When in total there are three persons online (the researcher, the other participant and you) it doesn’t matter who starts the conversation. You can start chatting by clicking on the chat-button in the right corner, see picture.
8. Remember that you can only ask questions of the list underneath. However, you could answer questions or ask the other participant to elaborate on a topic. However, you should not make up your own topic and stay close to the questions. 9. Go to the chatroom and chat (You must copy & paste the link in your browser): https://appear.in/onlinedatingexperimentuva
10. Fill in the personality questionnaire underneath.
11. Instruction: How well do the following statements describe the others’ personality? Put a ‘X’ on the place where you think it should be.
12. Save and send this document to the researcher.
13. Receive and open the file of the other participant that you get from the researcher.
14. Can you adjust the questionnaire in the file of the other participant with a ‘O’ on the place where you think your personality is? If it is on the correct spot already, you must place the additional ‘O’ next to it. It should look like this than ‘X O’.
15. How do you think the other participant described your personality in general (based off the personality questionnaire below)? Make your answer bold:
very inaccurately / quite inaccurately / neither inaccurately or accurately / quite accurately / very accurately
16. Save this document and send it to the researcher again.
17. The researcher tells you if you get interviewed first, or if you must wait for a maximum of 5 minutes. The researcher will send a message to you when you will get interviewed.
APPENDIX B: 36 QUESTIONS OF GROUP PERSONAL
1. Given the choice of anyone in the world, whom would you want as a dinner guest? 2. Would you like to be famous? In what way?
3. Before making a telephone call, do you ever rehearse what you are going to say? Why? 4. What would constitute a “perfect” day for you?
5. When did you last sing to yourself? To someone else?
6. If you were able to live to the age of 90 and retain either the mind or body of a 30-year-old for the last 60 years of your life, which would you want?
7. Do you have a secret hunch about how you will die?
8. Name three things you and your partner appear to have in common. 9. For what in your life do you feel most grateful?
10. If you could change anything about the way you were raised, what would it be? 11. Take four minutes and tell your partner your life story in as much detail as possible. 12. If you could wake up tomorrow having gained any one quality or ability, what would it be?
13. If a crystal ball could tell you the truth about yourself, your life, the future or anything else, what would you want to know? 14. Is there something that you’ve dreamed of doing for a long time? Why haven’t you done it?
15. What is the greatest accomplishment of your life? 16. What do you value most in a friendship? 17. What is your most treasured memory? 18. What is your most terrible memory?
19. If you knew that in one year you would die suddenly, would you change anything about the way you are now living? Why? 20. What does friendship mean to you?
21. What roles do love and affection play in your life?
22. Alternate sharing something you consider a positive characteristic of your partner. Share a total of five items. 23. How close and warm is your family? Do you feel your childhood was happier than most other people’s? 24. How do you feel about your relationship with your mother?
25. Make three true “we” statements each. For instance, “We are both in this room feeling ... “ 26. Complete this sentence: “I wish I had someone with whom I could share ... “
27. If you were going to become a close friend with your partner, please share what would be important for him or her to know. 28. Tell your partner what you like about them; be very honest this time, saying things that you might not say to someone you’ve just met.
29. Share with your partner an embarrassing moment in your life. 30. When did you last cry in front of another person? By yourself? 31. Tell your partner something that you like about them already. 32. What, if anything, is too serious to be joked about?
33. If you were to die this evening with no opportunity to communicate with anyone, what would you most regret not having told someone? Why haven’t you told them yet?
34. Your house, containing everything you own, catches fire. After saving your loved ones and pets, you have time to safely make a final dash to save any one item. What would it be? Why?
35. Of all the people in your family, whose death would you find most disturbing? Why?
36. Share a personal problem and ask your partner’s advice on how he or she might handle it. Also, ask your partner to reflect back to you how you seem to be feeling about the problem you have chosen.
APPENDIX C: 36 QUESTIONS OF GROUP FACTUAL
1. Where are you from?2. Are you having a good day? 3. Are you in college?
4. Do you have brothers or sisters? 5. Do you have any hobbies? 6. Are you single?
7. Do you like ice-cream? 8. What do you do for a living?
9. Are you a cat person or a dog person? 10. Are you a vegetarian?
11. Do you like scary movies? 12. What is your favorite drink? 13. Do you like tattoos?
14. Are you close to your family? 15. Do you play any instruments? 16. What is your favorite food? 17. Do you like reading?
18. Do you like to go to concerts? 19. Do you play any sports? 20. Do you like to go to the gym? 21. How old are you?
22. What for job do your parents have? 23. Which mobile phone do you have? 24. Do you have any pets?
25. Do you have roommates? How many? 26. Do you have any job on the side? 27. What’s your favorite TV show? 28. Anything you don’t like to eat? 29. Do you like gaming?
30. What kind of music do you like? 31. Do you like traveling?
32. Are you part of a student union? 33. When is your birthday? 34. What is your favorite color? 35. Do you like to go out? 36. Do you like shopping?
APPENDIX D: PERSONALITY QUESTIONNAIRE
Personality questionnaire
Disagree stronglyDisagree a little Neither agree nor disagree
Agree a little Agree strongly is reserved Is generally trustworthy tends to be lazy is relaxed, handles stress well has artistic interests is outgoing, sociable tends to find fault with others does a thorough job gets nervous easily has an active imagination
