Dissertation

In this study, three samples were collected to investigate the relationship between word choice and self-described personality traits measured by the Eysenck Personality Indicator (EPI). The data was analysed using SPSS and organised in Excel, with the Chi-square test and Fisher's exact test applied for analysis. While the null hypothesis was accepted, indicating no significant difference in word choice between introverted and extroverted individuals, the study showed potential for distinguishing between these word choices.

Hypothesis 1

• Word associations can be used to understand personality. Introverts and extroverts can be distinguished through word associations.

Prediction 1

• Extroverted individuals will use more sociable and bold language to describe themselves compared to introverted individuals even if they identify as neurotic or emotionally stable.

• Introverted individuals will use more reserved and reclusive language to describe themselves compared to extroverted individuals even if they identify as neurotic or emotionally stable.

Hypothesis 2

• Group dynamics change the word people use to describe themselves due to social cohesion altering people’s perception of themselves with ratios of sex and size of group altering group dynamics.

Prediction 2

• People will describe themselves differently after undertaking group tasks. As social cohesion may alter perception of oneself due to comfortability in groups possibly changing people’s perception of themselves as other people may interpret them differently.

Hypothesis 3

• People who feel more bonded to the group will have a greater change in personality. While those who haven’t despite extroversion/introversion will stay the same.

Prediction 3

• Introverts will be more set on the personality than extroverts when describing themselves as extroverts are influenced by sociability so their word choices might change.

• Group size alter levels of comfortability in group with those in smaller groups feel more included in the group dynamics; while those in larger groups will feel less included. This alters the perception of group dynamic possibly altering participants word choices as a result.

Hypothesis 4

• The ratio between males and females in group dynamics has an influence on how people interact and group comfortability.

Prediction 4

• Mixed gender groups will have different levels of inclusion to groups.

• Females will bond better with females than males; vice versa.

• Females will be more reserved with mixed sex groups and more assertive in female only groups.

Variables analysed in the chi-square test:

• Sex

• Extroverted/introverted?

• Emotional stability

• Eysenck description

• Self-described Eysenck extroverted/introverted ?

• Self-described Eysenck emotional stability

Variables not analysed in the chi-square test:

• Actual Eysenck’s results

• Word choices

Word Choices - Extroverted Sample

Experiment protocol

The experimental protocol was inspired by school games implemented in school.  Throughout the experimental measure, one moderator hosts the experiment session. This moderator note-takes the experiment session observing the icebreaker as well as tallying the game results.

So that the notes could observe any patterns in the participant’s behaviour that could have impacted the results as well as the general reception of the experimental session.

The test conditions of the experiment remained consistent in every experimental session to remain with the within-subject design by testing individuals in the same conditions to limit any possible biases in the parameters of the experiment.

The experimental protocols employed in the lab design included an icebreaker, a game of wink murder, a game of possibilities, and an experiment survey. Participants were instructed to share two truths and a lie about themselves and guess which of the answers was a lie or truth. Treats were distributed upon completion of the experiment survey.

Table 12 presented a slight pattern in results, with 5/6 of the individuals being introverted contesting the hypothesis that extroverts were more likely to attend experimental phases. The only emotionally unstable individuals had a change of self-described type from sanguine to phlegmatic but the same actual Eysenck type of melancholic. 

This suggests that the personality parameters of the EPI remained consistent and that the individual's perception was different due to emotional stability. Further research is needed to determine if this scale is suitable.

Was the hypothesis supported?

• The null hypothesis is accepted.

• The other hypothesis couldn’t be applied to this study due to issues with the pilot study sample

The main obstacles in the research process were in procuring enough participants and the loss of data from the second session. Content analysis was used instead of statistical analysis due to the meagre sample size and the difficulty of organising the combined data. Experimental Measures showed a null hypothesis due to a low sample size.

The findings concluded that there were some patterns in the introverted people being introverted, but many of the scores were skewed.

A new word choice  parameter was created , which had a significant influence on the research. Further research is needed to understand how language is impacted by socialisation alongside the EPI parameter to interpret the authentic influence on word choices and temperament. The null hypothesis was accepted due to a low sample size, but the patterns suggest that temperament can be influenced by comfortability. 

More research is needed to answer unanswered questions such as if temperament is innate, can it be influenced by society, and if personality parameters should be used to understand people. Understanding temperament provides the atmosphere to interpret socialisation in all forms, online and in person, which is essential in understanding what it means to be human.

The study was limited by time constraints and the questionnaire itself being too long. A self-perceived scale variable would be added to the study to interpret if participants felt closer to the self-described traits or the actual Eysenck type. 

The criteria in the study suggest a new way of exploring temperament, but it still leaves the biggest question: is temperament/personality influenced by word preferences? More cross-cultural emphasis is needed to truly understand language and its influences on personality. 

Suggesting for the personality parameter to test sociability would improve understanding whether pair-bonding is a great bonding and have an influence on at the language people use.

Thank you to the following people for helping me complete the project:

• My supervisor

• My colleagues

• Participants for my pilot study and questionnaire

• Stats clinic, Stats professors

• University of Kent, department of Anthropology & Conservation