Stranger interaction: factors that play roles in seating choice

COMM88

Fall 2018

TA: Sandi Moxley

University of California, Santa Barbara

 2

Stranger interaction: factors that play roles in seat choice

Since human beings are social groups, people are inevitable to enter an unfamiliar

environment in everyday life. In this case, the initiation of conversation and communication

with strangers is unavoidable. Why, in an unacquainted situation of free choice, does an

individual choose to sit close to and talk to a person instead of another person? What factors

are playing roles when people choose the person to sit close to? Is physical resemblance

affecting the seating choice in a room full of strangers? Series of experiments are performed

to investigate whether physical similarities have influence on seat choice when people enter

an unfamiliar environment.

Researchers have been interested in stranger interaction for a few years. A few

researches have been conducted to answer the questions above. However, most researchers

put emphasis on variables such as proximity and similarities in attitude and values rather than

physical similarities. One of the researches related is a study Mackinnon, Jordan, and Wilson

(2011) conducted to investigate how two physical similarities, glass wearing and sex, exert

effects on people’s choice of seat. In the study, seating arrangement was observed at a

computer lab with 31 seats in the library at a Canadian university. For each occupied seat in

the lab, the researcher recorded (a) the sex of the person sitting and (b) whether the person

was wearing glasses. Observations were recorded 1-3 times daily on weekdays, most

frequently during the midday, which was when the library computer lab was most populated.

The result shows that glasses wearers will tend to sit beside other glasses wearers more

frequently than expected by chance and women will tend to sit beside other women (and men

by men) more frequently than expected by chance. To verify the effect of sex, or say, gender

in a different environment, hypothesis 1 is below.

H1: Individuals are more frequent to sit close to people with the same gender.
 3

Besides gender, there are multiple aspects of similarity. Luo (2009) performed a

research to investigate the role of couple similarity in a sample of 117 couples who recently

started dating. During their visit to the lab, participants completed a packet that includes a

background questionnaire and a variety of inventories regarding their attitudes, values,

interests, and various personality characteristics. The result is that for interests, couples

showed a significant correlation on sport activity and habitual activity. In this case, it is

reasonable to assert that couples may show a significant correlation on body shape because

body shape, generally speaking, reflects sport activity and habitual activity. It is widely

acknowledged that approaching and conversation are first two steps of a relationship, which

leads to

H2: Individuals are more frequent to sit close to people with similar body shape.

Method

In order to examine the effect of gender and body shape when people choose seats in

an unacquainted environment, a quasi-experiment is conducted. A room which is capable to

contain 8-12 persons in the library of University of California, Santa Barbara is booked as

test field. The same chairs are put as a circle in the room. All chairs are divided into groups of

two and adjacent two chairs are in a group. All the distances between two adjacent groups of

chairs are similar. Participants do not know any volunteer who sits in the room. Participants

are asked to choose a chair to sit in the room and to complete a short questionnaire. In this

way, the real purpose of the experiment is concealed and personal information of participants

is collected. The seats they choose are recorded.

Sample

In the experiment, convenience sample is employed to collect data. Participants are randomly

selected from people who study in the library. They are asked whether they are interested in a

communication experiment or not to make sure they are voluntary to participate in the
 4

experiment. To examine hypothesis 1, a group of samples consists of 14 participants. 5 of

them are female, while 9 of them are male. To test hypothesis 2, two experiments are

conducted. The first experiment has 19 participants, among whom are 12 males and 7

females. The second experiment has 6 female participants.

Procedure

Before the experiments were performed, a pre-test was conducted. Among the twelve chairs

in the room, the six chairs which volunteers would sit later were not permitted to sit on.

Participants were free to choose any of the seat except for these six chairs. The seat each

participant chose was recorded. The result reveals that no chair is particularly popular. Every

chair has a similar frequency to be chosen.

In the experiment which investigates the influence of gender, all chairs in the room were

divided into six groups. Six volunteers, three males and three females, sit on chairs of six

different groups before each participant entered the room. Each seat had a vacant chair close

to it to make sure that each participant had no choice but to choose to sit next to one of the

volunteers. Each volunteer held several copies of questionnaire which contains basic

questions such as gender, height, and weight. All volunteers were similar in other physical

appearances except for gender. Participants were asked to choose a seat, gain a copy of

questionnaire from the volunteer next to them, and complete the questionnaire on the seat. In

first experiment which examines the effect of body shape, six volunteers are employed, three

males of different body shapes and three females of different body shapes. In the second

experiment, all volunteers are four males of different body shapes. All of other procedures are

the same as the gender experiment.

Measures

The dependent variable of the experiment is whether people sit close to people with same

gender and similar body shape. Frequency and percentage are appropriate tools to measure
 5

the dependent variable. After the participants finish the questionnaire and leave, who they sit

close to is recorded on the questionnaire. To quantify body shape, BMI (body mass index) is

utilized. According to National Heart, Lung, and Blood Institute, body mass index (BMI) is a

measure of body fat based on height and weight that applies to adult men and women.

To define what means similarity in body shape, the difference of BMI is used. BMI normally

has four categories – underweight, normal weight, overweight, and obesity. Although there

are some subtle differences between divisions of categories on internet, the categories of

normal weight and overweight tend to have range differences about 6. As a result, a

difference below 6 can be considered as an apt measure showing that a participant and a

volunteer are similar in body shape.

Results

Frequency and percentage are taken advantage of in the experiments to make

comparisons between different variables. H1 predicts that people are more likely to sit close

to people with the same gender. Analyses of frequency of how many individuals sit beside

people with same gender support this hypothesis. Table 1 demonstrates that 5 female

participants of 14 all sit close to female volunteers while 9 male participants all sit close to

male volunteers. The frequency is 14 and percentage is 100%, which is able to support the

H1. H2 predicts that people are more likely to sit close to people with similar body shape.

The first experiment discusses about BMI of same gender. Table 2 shows that 9 male

participants of all 12 males sit next to male volunteers whose BMI has a difference below 6

from them, which means 75% (9/12=0.75) of male participants choose seats next to people

with similar body shapes. It is also noticeable that among 7 female participants, 6 choose
 6

female volunteers who have similar body shapes, which leads to a percentage of 86%

(6/7=0.86). Therefore, in the case of same gender, H2 is supported. Whether body shape

effect has an influence in different genders is investigated in the second experiment.

According to table 3, 83% (5/6=0.83)female participants sit close to male volunteers who

have similar body shape with them. Hence in the case of different genders, H2 is supported as

well. Then, it is not difficult to draw the conclusion that H2 is supported no matter gender.

Discussion

The results illustrate that college students are more probable to sit close to people with

the same gender and similar body shape when they enter an unacquainted environment. All

these are consistent with the study of Mackinnon, Jordan, and Wilson and Luo. Due to the

characteristic of experiment, the external validity is not satisfactory enough, which means

that these data are of limited generalizability. This may be true if compared with surveys.

However, if so, it is equally true for most other studies. Data in this experiment are collected

in an actual field setting – a library, which is one of the scenes of everyday life. Compared to

the results of other studies that are performed in entirely experimental environments, these

results have higher external validity because the actual field setting enables naturally, instead

of artificially, free choice to occur.

The largest limitation of these data, however, lies in internal validity. Because of the

restriction of time and location, convenience sample has to be employed. All participants are

college students who study or have a rest in the library. When selection of subjects is not

really a random sample, selection bias exists. In such case, one of the most significant parts

of true experiment is lack – random assignment, which is also limitation of majority of quasi-

experiments. This is the reason why this experiment cannot contribute to causality. Instead,

the only conclusion researchers can draw is that gender and body shape of volunteers have

strong relationship with seat choice of participants. To achieve a causal conclusion, richer
 7

data of different groups of people in the society are essential. The other important element of

internal validity is to control the environment well in order to eliminate other explanations. In

the first experiment to examine BMI effect, as a result of time restriction, three male

volunteers of different body shapes and three female volunteers of different body shapes have

to be in the room at the same time, which means, as a matter of fact, two independent

variables are included in this round. This can cause an extreme descent in internal validity.

Nonetheless, if the result of former experiment that 100% participants sit next to volunteers

of the same gender is taken into consideration, the serious consequence of result error can be

evited since gender diversity will not affect the choice of BMI. Hence, the result can still be

persuasive and compelling. However, for the sake of safety, this experiment is supposed to be

divided into two sub-experiments, one with 3 male volunteers of different body shapes and

the other with 3 female volunteers of different body shapes.

In consideration that it is a pre-test/post-test experiment, the potential problem that

almost all pre-test/post-test experiments have also unavoidably exists in this experiment. The

problem is that differences in data of dependent variables could be the result of interaction of

manipulation with pre-test. To be more specific, when participants enter the room for the

second time, they are likely to sit on the same chair they have sit before regardless of who sits

beside them. It is understandable because human beings tend to be afraid of unfamiliar

environment and be more comfortable in where they are familiar. Due to the lack of Solomon

4-group design, the comparison of pre-test/post-test group and post-test only group cannot be

made so it is difficult to determine whether pre-test has an effect on seat choice. To solve this

problem, Solomon 4-group design is necessary and more time is needed.

The experiment also contains a questionnaire to collect the personal information of all

participants. The questionnaire is consisted of three factual question, including gender,

height, and weight, all of which are open-ended questions. All the questions are succinct and
 8

pellucid to make sure that there are no other interpretations of questions. Under this

circumstance, all the information collected is self-reported. Therefore, it is reasonable to

doubt the authenticity of information because factually inaccurate response may be gathered.

It is likely that participants do not know their accurate height and weight and report an

approximate number. Or the height and weight they know have changed and are different

from the current data. The result of BMI calculation is likely to be impacted. As a result, error

may exist in the frequency of body shape group.

Clearly, other physical traits such as hair color and race need to be examined to more

adequately. More data of samples are supposed to be gathered and the experiment should be

replicated to achieve higher reliability. In addition to the methodical problems stated above,

there are some other social and practical issues. In the contemporary society, gender cannot

be briefly categorized as male and female. A few individuals have gender identity and

expression that does not align with male and female. Those people are ignored automatically

in this experiment. Another problem is that sometimes subtle difference in BMI cannot show

evident difference in body shape. From table 2, one male volunteer has BMI of 18.4 and one

male volunteer has BMI of 19.1. These two numbers are close and the difference is hard to be

told if participants just use their eyes to make a decision. It is difficult for researchers to find

volunteers whose body shapes are broadly different. But because the experiment investigates

similarity, the impact of two volunteers with similar body shape can be omitted. From table 2,

BMI range of male volunteers is 10 (28.4-18.4=10), while that of female volunteers is 9

(26.2-17.2=9). Because the range of BMI is large enough, at least body shapes of two

volunteers have apparent difference, the conclusion can still be drawn.

One of advantages of this experiment is that it possesses high objective validity,

which means it can predict the outcome of a similar situation based on the study findings.

When a college student enters a room full of strangers, where he or she tends to sit and who
 9

he or she tends to initiate a conversation with can be predicted. This strategy can be applied

to considerable fields such as advertising and marketing. Future studies can further

investigate the topic of seating choice. Will people with physical resemblances continue to

talk and talk more? Will they develop their relationship more rapidly than those with rare

physical resemblances? Are they conscious or unconscious to make such choice? All those

are waiting to be answered in the future.

 10

References

Calculate Your BMI - Standard BMI Calculator. (n.d.). Retrieved from

https://www.nhlbi.nih.gov/health/educational/lose_wt/BMI/bmicalc.htm

Luo, S. (2009). Partner selection and relationship satisfaction in early dating couples: The

role of couple similarity. Personality and Individual Differences, 47(2), 133-138.

doi:10.1016/j.paid.2009.02.012

Mackinnon, S. P., Jordan, C. H., & Wilson, A. E. (2011). Birds of a Feather Sit Together:

Physical Similarity Predicts Seating Choice. Personality and Social Psychology

Bulletin, 37(7), 879-892. doi:10.1177/0146167211402094

 11

Tables and Graphs

Table 1

gender of participants gender of volunteers

0 0 　
0 0 　
0 0 　
0 0 　
0 0 　
1 1 　
1 1 　
1 1 　
1 1 　
1 1 　
1 1 　
1 1 　
1 1 　
1 1 　
(0 means female. 1 means male.)

Table 2

volunteer & BMI of gender of difference

gender Volunteer participant BMI of P Participant of BMI

1M 18.4 1 21.1 M 2.7
2 18.4 M 0
3 22.9 M 4.5
4 22.3 M 3.9
2M 28.4 5 22.3 M 6.1
6 25.6 M 2.8
7 20.6 M 7.8
8 23.1 M 5.3
3M 19.1 9 23 M 3.9
10 32 M 12.9
11 19.8 M 0.7
12 23.1 M 4
4F 26.2 13 19.3 F 6.9
14 20.5 F 5.7
5F 17.2 15 17.6 F 0.4
16 21.9 F 4.7
17 17.6 F 0.4
18 17.9 F 0.7
6F 20.2 19 18.6 F 1.6
 12

Table 3

gender of BMI of BMI of BMI gender of

participants participants volunteers difference volunteer

F 22.3 18.4 3.9 m
F 19.3 26.2 6.9 m
F 20.5 26.2 5.7 m
F 19.8 19.1 0.7 m
F 23.1 28.4 5.3 m
F 18.4 18.4 0 m
 13

Appendix

Questions on the questionnaire.

1. What is your gender?

2. What is your What height?

3. What is your weight?

4. Please draw a cute picture below.

Abstracts

Birds of a Feather Sit Together: Physical Similarity Predicts Seating Choice

Across four studies, people sat (or reported they would sit) closer to physically similar

others. Study 1 revealed significant aggregation in seating patterns on two easily observed

characteristics: glasses wearing and sex. Study 2 replicated this finding with a wider variety

of physical traits: race, sex, glasses wearing, hair length, and hair color. The overall tendency

for people to sit beside physically similar others remained significant when controlling for

sex and race, suggesting people aggregate on physical dimensions other than broad social

categories. Study 3 conceptually replicated these results in a laboratory setting. The more

physically similar participants were to a confederate, the closer they sat before an anticipated

interaction when controlling for sex, race, and attractiveness similarity. In Study 4, overall

physical similarity and glasses wearing similarity predicted self-reported seating distance.

These effects were mediated by perceived attitudinal similarity. Liking and inferred

acceptance also received support as mediators for glasses wearing similarity.

Partner selection and relationship satisfaction in early dating couples:

The role of couple similarity

Although much research has been conducted to test the role of couple similarity in

partner selection and predicting relationship satisfaction, previous research was limited to
 14

couples in well-established relationships. The current study examined couple similarity in a

sample of 117 couples who recently started dating (dating length 6 six months). The early

dating couples showed strong to moderate similarity on demographics (age and ethnicity),

personal interests, values, and political attitudes but weaker similarity on a range of

personality domains. With regard to satisfaction prediction, the absolute difference tended to

be a better predictor than the profile similarity. However, couple similarity overall was not a

strong predictor.