Investigating the link between subjects introversion/ extroversion scores, and their differing abilities in performing the same task, alone and with an audience.

By Duncan McPherson (PS1002)
5th of April 1996

Abstract

This study is concerned with examining the correlation between subjects introversion/ extroversion "E" scores (as measured by the Eysenck Personality Questionnaire) and their differing abilities in performing the same task in front of an audience, and alone. The study draws on theories of Central Nervous System arousal by Eysenck (1975 & 1983) , Zajonc (1965) and Yerkes - Dodson.

The task given was a drawing task where subjects could only see what they were doing via a mirror. Each subjects ability to perform the task (accuracy and time taken) was recorded, under the two conditions :

i) The subject performing the task alone

ii) The subject performing the task in front of an audience

A single performance score was produced for each subject indicating which condition was performed best at. The E scores and the performance scores for all subjects were then correlated.

This study found that there was no significant correlation between E scores and Performance scores.

Introduction

This study is concerned with investigating any link between subjects differing abilities to perform a task in front of people and alone, and their introversion/extroversion scores, as measured by a personality test.

The work of both Eysenck and Eysenck (1975 & 1983) and Zajonc (1965) has produced hypothesis relevant to this study. Eysenck states that introvert subjects have a higher natural level of Central Nervous System (CNS) "arousal" (also known as "drive"). Zajonc makes the further claim that the presence of other people will increase any individuals level of CNS arousal. These two hypothesis can then be combined with the Yerkes - Dodson law. This states that every task will have an optimum level of CNS arousal. That level will differ depending on the type of task. If a subjects level of arousal is above or below the optimum level then then their performance in that task will deteriorate. Only when the subjects level of CNS arousal is "just right" for a certain task, will he or she perform that task to the best of their abilities.

Graph 1

Assuming that Eysenck and Zajonc are correct, increasing a subjects level of arousal should have a different effect on task performance depending on whether that subject is an introvert or an extrovert. This is due to differing "base levels" of arousal in introverts and extroverts, and the change in task performance if their arousal moves towards or away from the optimum.

Graph 2

As an example to illustrate this theory, graph 2 shows a task performance curve for a task which is optimally suited to an unaroused introvert subject, i.e. the subject is alone and at their usual "baseline" level of arousal. By comparison an extrovert subject with the same abilities as the introvert should be less aroused when alone. This would lead to the introvert being better at the task than the extrovert. However when both introvert, and extrovert are required to perform the same task with others watching, the situation changes. Both subjects level of arousal will be increased (according to Zajonc), causing the introverts arousal to become higher than the optimum level for the task, whereas the extroverts will increase to that optimum level. The net result should be a marked increase in the extroverts performance and a similar decrease in the introverts performance. Changes will of course depend on each individuals "baseline" level of arousal and the type of task involved. Harder tasks have a lower optimum level of arousal than easier tasks. As a consequence of these theories of arousal Eysenck suggests that introverts and extroverts will be better at different types of tasks in everyday life. Increased levels of arousal supposedly increase individuals tendency to make dominant (automatic) responses. High levels of arousal would then only be conducive to easy tasks. If this is the case, it suggests that monotonous tasks may be better suited to introverts due to their naturally high levels of arousal.

The task performance of each subject can be relatively easily measured (e.g. by the number of errors made and the time taken to complete), however measuring a subjects introversion/ extroversion is a more complex issue. In this study the Eysenck Personality Questionnaire (EPQ) has been used. This reveals three dimensions of a subjects personality, each dimension measured on a scale of 1 to 12. The dimensions are as follows :

E score= "Introversion"/"Extroversion" scale (1 is most introvert and 12 is most extrovert).

P score = "Psychotocism" (1 being most psychotic). In the case of the EPQ "psychotic" is defined as solitary people, not caring about or fitting in well with people.

N score = "Neurotocism" scale (12 being most neurotic). The EPQ defines neurotic as typically anxious, moody and over emotional people (also prone to depression). Those scoring low N scores are usually calm and unworried by life.

In this study only the introversion/extroversion, "E" scores from the subjects EPQ are used. There is also a fourth "lie" scale present in the EPQ, which if the subject scores too highly in renders the test invalid. The lie scale is designed to measure subjects tendency to "fake good". Each subject will also have a score in task performance, with every subject performing the given task alone and in front of a group of people. This will be more negative if the subject is better at doing the task alone, and more positive if the subject performs better with and audience. Finally both E and performance scores for every subject will be statistically analysed for any correlation.

Hypothesis

Null Hypothesis (H0) : There will be no significant correlation between introversion/ extroversion scores on the EPQ, and differing abilities in solitary performance of a task and performance in the presence of others.

Alternative Hypothesis (H1) : There will be a significant correlation between introversion/ extroversion scores on the EPQ, and differing abilities in solitary performance of a task and performance in the presence of others.

Method

Subjects

The subjects used in this study were selected from the 1st year psychology practical group undertaking the study,i.e. by availability sampling. There were 28 subjects used in the study with a mode age of about 18 - 19. All were psychology students.

Materials

The study made use of the short scale Eysenck Personality Questionnaire devised by Eysenck and Eysenck (1975 &1983). This consisted of 48 true/false typed questions designed to identify personality traits in individuals. Also used were transparent marking grids to allow subjects to assess their own EPQ scores on the four personality scales measured. Each subject was given a task sheet with two star outline sets photocopied on it (one for each condition).

Apparatus

The task given was to draw a line between the outline of one of the "stars" on the task sheet (see diagram). To make this a difficult task subjects were unable to view the outlines they were drawing between directly. A clamp stand with a piece of opaque card attached was put in the subjects line of sight preventing them from seeing, except via a mirror. A stopclock was used to determine the time the subject had taken to draw a complete outline of one star.

Design

The independent variable was whether the subjects were required to perform the task alone or in the presence of other subjects. The correlational (dependent) variables measured in this study were each subjects introversion/extroversion "E" scores, and a score of their performance in the task. The task performance score was a single value for each subject, more negative if the subjects performance was better alone and more positive if it was better with an audience. The experiment used a counterbalancing design to attempt to eliminate any learning effect that may have occurred. To achieve this approximately half of the subjects performed the task in front of an audience first, the rest doing the task alone first. The study used a within subjects design with all subjects performing the task under both conditions.

Procedure

Each Subject firstly completed an Eysenck Personality Questionnaire (EPQ), afterwards assessing their own score using the marking grids provided. The E scores were noted. Subjects were then each given a task sheet and split into groups of approximately six. Half of the groups performed the task individually first, each member in a separate room (alone condition). The other half performed the task one at a time in front of the other members of that group (audience condition). Using the apparatus described, each subject "drew" inside the outline of one of the stars on their task sheet. The number of mistakes and the time taken to draw all the way round the star was noted for each subject. Finally all groups swapped over (counterbalancing) and performed the task under the condition they had not yet done.

Results

For each subject two sets of data were noted, task performance and E score. The task performance was calculated as follows for each subject :

A = Time taken to draw inside outline of star in alone condition

E1 = Number of mistakes made in alone condition

B = Time taken to draw inside outline of star in audience condition

E2 = Number of mistakes made in audience condition

Task Performance = ( A + E1 ) - ( B + E2 )

Once task performance scores and E scores for all subjects were collected the data was correlated using a spearman rank correlation. The data is shown as a scattergram in graph 3. The spearmans rank correlation was found to be 0.19. This correlation is not significant at the 0.05 level with 28 subjects (see table 1). Thus we can reject the alternative hypothesis (H1) and accept the null hypothesis (H0) :

There will be no significant correlation between introversion/extroversion scores on the EPQ, and differing abilities in solitary performance of a task and performance in the presence of others.

Table 1

Graph 3

Discussion

After statistically analysing the results from the experiment it was found that there was no significant correlation present. Therefore this study does not provide any support for the conclusions of Zajonc, Eysenck or Yerkes - Dodson. Other studies do support the link between introversion/extroversion and CNS arousal. Work by Harkins and Green (1975) found that introverts are less easily distractable than extroverts. However conclusions drawn by Eysenck and other psychologists supporting his theories of personality, are still under dispute in the scientific world.

Possible Sources of Experimental Error

The use of the EPQ in this study had one major shortcoming. The test has a built in "lie" scale, which if the testee scores too highly in, means that the test is invalid. In this study the lie scale data was ignored. This could have seriously damaged the credability of results in the experiment as invalid intorversion/extroversion scores may have been used in the final correlation.

The EPQ was given to all the subjects in the same room at the same time, where subjects may have taken too long to "think". A social situation such as this may have had an effect on the outcome of the EPQ. One possible improvement could have been to conduct EPQ tests under more stringent "exam" conditions.

The overall validity of the EPQ in determining whether subjects are introverts or extroverts can also be questioned. As with all personality tests it is very difficult to verify the EPQ with other data, and thus decide the validity of what the test claims to measure. The method of factor analysis used to produce the EPQ is also under question is psychology, as there can be no guarantee that the traits picked up in this way are "real" in the testee's personality. They could be merely traits in how we answer personality tests, that have been picked out by the factor analysis.

The task given may have been novel to all the subjects in its specifics, however some subjects may be naturally better at drawing tasks such as this. It could be the case that lots of previous drawing experience made a subject "worse" at the task, because he/she needed to "re-learn" how to draw looking into a mirror. This shortcoming could be solved by devising more tasks and averaging each subjects performance over several of them.

Also with respect to the counterbalencing used in this study. The mirror drawing task may have had a substantial learning effect attatched in some subjects only. One way to alleviate this could be to use a larger selection of subjects. This would help to "spread out" subjects who learned the task quicker.

Finally this study was conducted using Psychology students as the subject pool. This availability sampling greatly reduces the external validity of the results. In order to make more general conclusions this study should be replicated with a larger population of people from different backgrounds (random sampling).

References

Gleitman, H. (1991). Psychology. University of Pennyslvania : W.W.Norton and Co.

(pp 510-511)

Eysenck, H. J. and Eysenck S. B. G. (1975) Psychotocism as a dimension of personality. London : Hodder and Stoughton.

Eysenck, H. J. and Eysenck S. B. G. (1983). Recent advances : The cross cultural study of personality. In Butcher, J. N., and Spielberger, C.D (Eds), Advances in personality assessment, vol. 2, pp 41-72. Hillsdale, N.J.: Erlbaum

Zajonc R.B. (1965). Social facilitation. Science 149:269-74


Duncan placed this report online on the 11th of March 1997
For more information contact dfmcp@hotmail.com