A Comparison of Alternative Measures of Individual Differences in Brain Lateralization

Citation:

Flemming Hansen and Niels Erik Lundsgaard (1982) ,"A Comparison of Alternative Measures of Individual Differences in Brain Lateralization", in NA - Advances in Consumer Research Volume 09, eds. Andrew Mitchell, Ann Abor, MI : Association for Consumer Research, Pages: 161-169.

Advances in Consumer Research Volume 9, 1982      Pages 161-169

A COMPARISON OF ALTERNATIVE MEASURES Of INDIVIDUAL DIFFERENCES IN BRAIN LATERALIZATION

Flemming Hansen

Niels Erik Lundsgaard

ABSTRACT -

Four different measures of individual differences in brain lateralization have been intervalidated; i.e. tachistoscope measures, dichotic listening, selected psychological tests, and self-administered questionnaires. Two times 50 respondents were exposed to the various stimuli and factor analysis has been used to analyze the collected data. Also part of the data have been reliability-tested.

The present findings suggest that there are complicated problems involved in measuring brain lateralization and recommend that future efforts are concentrated on methodological problems, rather than wasted on speculative applications, resulting in findings the interpretation of which are far from obvious.

Consumer behavior researchers have shown an increasing interest in brain lateralization in recent years. Based upon psychological studies of hospitalized patients, it became established in the early sixties, that the left and the right hemispheres of the human brain are functioning very differently. Studies of "split-brain patients" [As a treatment in severe cases of epilepsy, Bogen (1977) began a treatment, where the two brain halves were disconnected by having the corpus callosum (the nerve tissue connecting the two brain halves) cut. These "split-brain patients" made it possible to study the functioning of the left and the right brain half separately.] (Sperry, 1973) and another group of patients, to whom electric stimulation is applied to only one of the halves of the scalp (Deglin, 1976) as well as studies of individuals with partly damaged brains, all show the same. The left part of the brain is particularly specialized in handling verbal, symbolic, arithmetic information, and it is doing so in an orderly sequential manner. The right brain on the contrast, is concerned with holistic, musical, geometrical, and pictorial information. Moreover, the processes of the right brain are largely unconscious.

These observations have made researchers, in several areas, interested in possible applications. Wexler (1980) for instance, reviews left and right brain processing in relation to psychiatric treatment. Bogen (1977) discusses educational implications, and Hansen (1981) reviews findings of relevance for the study of individual information processing.

Traditional models of consumer behavior are decision process models, cognitive models, information processing models, and effect hierarchy models. The processes described in these models are so, that it is natural to suggest that they primarily occur in the left hemisphere. The question then arises whether there are right brain processes, also, with which the student of consumer behavior should be concerned ? To explore this possibility, it is necessary to discuss two alternative ways of studying brain lateralization.

The first approach focuses on whether different kind of stimuli tend to be processed more or less consistently in left or right brain hemisphere, and on whether different situations give raise to more or less right or left brain processing. Thus, the concern is with the extent to which certain brain functions are more likely to become activated under same circumstances, whereas other functions are triggered by other tasks or conditions. This, for example, was the approach taken by early psychologists studying split-brain patients. Here the concern was with the extent to which verbal vs. pictorial information gave raise to right or to left brain processing; how musical impressions were received, how arithmetic problems were handled, etc. Similarly studies where the left or the right brain half was neutralized for a shorter period of time with use of electro-shock were carried out along these lines. Later, also, many studies of normal individuals' perceptual kinds of stimulations gave raise to left or to right brain actiVitY (Kimura. 1973).

Also, in the study of consumer behavior the possibility of situational differences has been suggested. Krugman (1977) maintains that compared with print advertising, television advertising generates right brain information processing, and therefore, it functions very differently from what is suggested by effect hierarchy models.

In this connection Weinstein, Weinstein, and Appel (1980), and Appel, Weinstein, and Weinstein(1979) have reported relevant findings. With the use of electroencephalogram (EEG) measures, they studied left and right brain processing in adult subjects, while they were watching television commercials and print advertisements. Their findings, however, are not very conclusive and even though Krugman (1980) by reanalyzing their data gets somewhat more confirmative results, there are still many problems associated with the use of the EEG-measures in studies of brain lateralization. Presumably the measures are very sensitive to the location of the electrodes on the skull of the respondent, and it is also difficult to interpret the continuous stream of responses. Questions are such as: How long time intervals to use as analytical units? Should brain activity in a particular moment be related to the stimulation received simultaneously or to the information received sometimes earlier?

Rockey, Greene, and Perold (1980) also report on problems with the interpretation of brain wave-measures. In an experiment where they compare 8 different EEG-lateralization measures for advertisements known to produce high and low recall scores, and regular TV-programs, they get significant, but sometimes unpredicted different scores when viewing of programs and of commercials are compared. When high and low recall advertisements are compared, the resulting scores are even more difficult to interpret.

Finally, the use made by Kling (1980) of conjugate lateral eye-movements should be mentioned. With this measure, it is assumed that a person when faced with a task will look to the right, if he is primarily using the left brain,and that he looks to the left, if he is dominated by right brain processing. With this technique Kling finds some relationships between his respondents' tendency to rely upon left brain processing, and the consistency of judgements regarding car preferences.

Following the second approach the researchers look for stable individual differences in the extent to which people tend to rely upon left vs. right brain information processing. This approach has dominated among psychiatrists (Wexler, 1980) and among educational psychologists (Wittrock, 1977). Similarly, Richardson's (1977) work on visualizers vs. verbalizers as information processing types is relevant in this context. Also, it can be mentioned that Hansen and Lundsgaard (19813 attempted to develop operational measures for the purpose of identifying individual differences in brain lateralization, and it is remarkable that Appel, Weinstein, and Weinstein (1979), without emphasizing themselves, find dramatic individual differences in their brain lateralization index among their 31 Connecticut housewife subjects (see table 10 in their original paper).

The present paper looks at alternative ways of identifying such individual differences. To the extent that such differences can be established, they can explain individual differences in information processing, decision-making and persuability in connection with different kinds of messages.

MEASUREMENT PROBLEMS

There are several different techniques which can be used for the purpose of identifying individual differences in the extent to which individuals rely more or less extensively on left or right hemispheres. Several of these are reviewed in Hansen (1981). Among these the most promising are:

1. Tachistoscope measures

2. Dichotic listening

3. Selected psychological tests

4. Self-administered questionnaires

5. Electroencephalogram (EEG)

6. Conjugate lateral eye-movements (CLEM)

To cross validate these techniques is an extremely important task, which must be completed before a thorough understanding of individual differences in brain lateralization can be gained. Without this kind of validation it is difficult to compare findings from different research traditions and to improve our understanding of the nature of individual differences in brain lateralization research. Strangely enough, in spite of the frequent use of all of the techniques by different researchers, no attempts to cross validate any two of the techniques have been published. In the present paper, the four first of the techniques are studied. It is hoped that it shall subsequently be possible to extend the work so that EEG- and CLEM-measures will be included in the comparisons, also.

RESEARCH DESIGN

In the present study 50, 20-24 years old male and female, right handed, students in an educational program participated as subjects. Each subject was exposed to a tachistoscope test, a dichotic listening task, a selection of psychological tests, and finally they completed a self-administered questionnaire dealing with possible aspects of brain lateralization. Because of incomplete data 4 of the subjects subsequently were deleted, so that 46 complete data sets resulted.

Since some of the measurements are rather complicated,they will be described in more details, as they are introduced in the following.

For their participation respondents were paid a fee of $10.00 each. For each respondent the complete test took 2-3 hours. The data collection was carried out under supervision of the authors, and the psychological tests were administered by a trained psychologist, whereas the tachistoscope and dichotic listening tasks were carried out by a research assistant, especially trained with this equipment.

Tachistoscope Test

[The authors are grateful to associate professor Steen Larsen, Institute for Educational Psychology, Teachers College, Copenhagen, for assistance with the equipment for these tests.]

The tachistoscope test relies on that peculiarity of the human vision, that subjects appearing in the left visual field, is initially transmitted to the right brain half, and vice versa. It is important to note that we are talking about visual field and not right vs. left eye. Operationally this implies that it is necessary to have the respondent fixating his attention on a particular point, before information is exposed to the left and to the right of this point. For the present tachistoscope task, a design was used where a screen was exposed from behind. On the center of the screen, a circle is located, and at the beginning of the test, the respondent is asked to focus on this particular circle. Then, a figure or a letter is presented in the circle, together with the test material, which is exposed to the left and to the right of this point of fixation. Following this, the respondent is then asked to report what the figure or the letter in the circle was. This is done to make sure that the subject had actually had his attention directed to the center part of the screen.

One problem with the tachistoscope procedure is that even though information initially is transmitted as described from one visual field to the opposite brain half, it is perfectly possible for normal individuals, shortly after the initial exposure, to transmit the information back to the other brain half. Actually, when the information has first been transmitted to the right brain half, and you ask the respondent to verbalize what he has seen, he must utilize his left brain half to be able to verbalize his response.

Two different tests were used. One was supposed to favor the right brain half information processing, and the other was supposed to favor left brain half information processing. The first test consisted of showing circles in different locations in the left and in right visual field. One particular presentation looks as in figure 1A. Here an "A" appears in the center circle, and there are two other circles one to the left and one to the right of the focal point. After a brief exposure, the respondent is asked to indicate on a piece of paper, where he saw the circles. This is done by marking a sheet like figure 1B. This complete task consists of 34 exposures, and 2 initial ones where the respondent is instructed about the procedure. The exposure is very short (5-20 milli-seconds). With exposures of this length, it was determined in advance that normal individuals should be able correctly to identify approx. 75 percent of the circles exposed to them

FIGURE 1

BILATERAL CIRCLES (A), AND SCORING SHEET (B)

This as well as the next test was introduced to the respondents as a test of visual ability.

In the second test the respondents are exposed to three digit numbers in the left and in the right visual field. After the exposure which was 100-300 milli-seconds - which is longer than in the first test - the respondent was asked to identify one of the figures exposed, either to the left or to the right.

Whether the respondent was asked about the figure appearing in the left or the one in the right visual field was determined in advance in a manner, so that prior to the exposures, the respondent did not know what visual field he was to focus on. In this test, the score was computed as the number of correctly identified left or rightly exposed figures. In this fashion four different scores were obtained:

I. A tachistoscope circle -

    left visual field.................................(TA-C-L)

II. A tachistoscope circle -

     right visual field .............................(TA-C-R)

III. A tachistoscope figure -

      left visual field ..............................(TA-F-L)

IV. A tachistoscope figure -

      right visual field ............................(TA-F-R)

Since each of those scores are influenced not only by the individual 's tendency to rely more or less on left or right brain processing, but also by the individual's overall ability to identify circles or to identify figures, it can be useful to compute an index also neutralizing this effect. This index is computed as shown below, with the right visual field scores less the left visual field scores divided with the sum of the two scores:

V.   TA-C-IN = (TA-C-R) - (TA-C-L)

                           (TA-C-R + (TA-C-L)

VI.   TA-F-IN = (TA-F-R) - (TA-F-L)

                            (TA-F-R) + (TA-F-L)

Finally, a special score was computed based on both of the tests. Since the circle test was thought to favor right brain dominated individuals, and the figure test left brain dominated individuals. a score like:

VII.   (TA-C-L) - (TA-F-R)

might reflect right or left brain dominance.

DICHOTIC LISTENING

The dichotic listening task relies on a phenomenon which 55 somewhat similar to that based upon which the tachistoscope test is developed. It is known that the majority of the information sent to the left ear initially is processed in the right brain half and vice versa. Consequently, by sending competing information to the left and to the right ear, and by observing whether the individual picks more or less information up with his left or with his right ear, one gets an indication of brain lateralization.

For the present test, a specially developed stereo equipment was used. This was constructed so that with the use of ear-phones, one sequence of messages is sent to the left ear simultaneously with another sequence being sent to the right ear. The pitch of the messages sent to the left and to the right ear was adjusted, so that it would seem the same to normal individuals. The first test was composed of 30 meaningless three letter words. These words were presented in blocks of three, so that the respondent first heard two times three words with half of them sent to the left and the other half to the right ear. Each pair of words were sent at exactly the same time. Examples of the words are: GUK-BAP, VEP-GIT, B0T-VYK.

After the exposure to such two times three words, respondents were asked to report what they recalled. Again the score is a simple count of the number of correctly identified words sent to the left and to the right ear.

Following this task, the test was repeated with 30 three letter meaningful Danish words, with the scoring done in the same manner. Whereas the meaningless words were supposed to favor the right hemisphere processing, the meaningful words were supposed to favor left brain activity.

Out of the complete task came four basic scores. namelY:

I. Dichotic meaningless words

    picked up with left ear.................... (DI-ML-L)

II. Dichotic meaningless words

     picked up with right ear .................(DI-ML-R)

III. Dichotic meaningful words

      picked up with left ear ..................(DI-MF-L)

IV. Dichotic meaningful words

      picked up with right ear................ (DI-MF-R)

As with the tachistoscope test, an index was computed for the meaningful words, as well as for the meaningless words:

V.   D-ML-IN = (DI-ML-R) - (DI-ML-L)

                            (DI-ML-R) + (DI-ML-L)

VI.   D-MF-IN = (DI-MF-R) - (DI-MF-L)

                     x        (DI-MF-R) + (DI-MF-L)

Finally, a combined score was computed as:

VII.  (DI-ML-L) - (DI-MF-R)

As with the tachistoscope test, there are uncertainties in connection with the interpretation of the dichotic listening test results. Again here, information may very shortly after it has been received be transmitted to the opposite brain half, and to verbalize the response will always require the left brain half to become activated regardless of what ear the information originally was transmitted to. The dichotic listening test is complicated to construct. However, since the problems in connection with this test are discussed elsewhere (Lundsgaard and Hansen, 1981) they will not be taken up in any great details here.

It should be mentioned, however, that since it was possible to carry out a re-test with 20 of the original SO respondents, a reliability coefficient could be computed. With 3-4 weeks between the first and the second test, it was .6. Although not extremely high, it does suggest that the measurement procedure has some reliability over time. A finding also Wexler and Heninger (1979) report from a study with a very similar procedure.

TENDENCIES TOWARDS BRAIN LATERALIZATION AS MEASURED WITH TACHISTOSCOPE AND DICHOTIC LISTENING

Since each of the two tachistoscope tests and the two dichotic listening tests produce results with very different average number of correctly identified test items, and since also the variance differ in the four sets of measurement, it was decided to standardize and normalize the scores before further analyses were carried out. This was done by deducting the average from each single score and dividing the score with the variance.

After this all 8 basic measurements were correlated with each other. Results from this analysis are shown in table

TABLE 1

CORRELATIONS BETWEEN 4 TACHISTOSCOPE AND 4 DICHOTIC RAW-SCORES.

It is remarkable that none of the correlations are very high. Actually, the largest correlation (0.47) is between the measurement of correctly identified circles in the left visual field (TA-C-L) and correctly identified circles in the right visual field (TA-C-R). Presumably, respondents, who are good at identifying circles in their left visual field, are good at doing so in their right visual field, also. A similar phenomenon does not occur for the three other tasks for which measurements are made.

The next two observations to make from table 1, are the significant correlations between:

1. Tachistoscope circle left visual field measure (TA-C-L) and the measure for dichotic meaningless left ear words (DI-ML-L) (0.35), and

2. The similarly large correlations (0.42) between identification of tachistoscope circles in the right visual field (TA-C-R) and correctly identified meaningless words from the dichotic listening task (DI-ML-R) with the right ear.

Seemingly, the tachistoscope circle test and the dichotic meaningless word test are measuring a related phenomenon. This can hardly be said about any of the other measures, even though a few meaningful relationships approaching significance, appear in the table. For example, this is the case with the correlation between tachistoscope figures left visual field (TA-F-L), and dichotic meaningless left ear (DI-ML-L); the correlation between dichotic meaningless left ear (DI-ML-L) and dichotic meaningful left ear (DI-MF-L) and the correlation between the tachistoscope circles right visual field (TA-C-R), and tachistoscope figures right visual field (TA-F-R).

A similar correlation occurs, however, between tachistoscope circles, right visual field (TA-C-R) and dichotic meaningless words in the left ear (DI-ML-L). This may follow, however, from the high correlation between the left and the right measures for the tachistoscope circle task. Since these two variables are so relatively closely correlated, it is likely that anything correlated with one of them, is likely to be correlated with the other also.

Finally, it is note-worthy also, that 6 out of the negative correlations are between left and right brain measures, and that all measures supposed to measure right brain dominance, are positively correlated and so are all but 2 measures of left brain dominance.

Altogether the correlations of table 1 are not very promising even though they do suggest a possibility for continued work with the tachistoscope circle test combined with the dichotic meaningless word test. This is explored further in a different context. In the present paper, focus is on the validation problems in connection with the four different measures studied.

The close relationship between the measures of correctly identified circles in the left (TA-C-L) and in the right hemisphere (TA-C-R) makes it somewhat difficult to interpret the only strongly significant relationship found in table 1. This suggests that the indexes computed as described before, may be useful in identifying tendencies for the left vs. the right brain half to dominate in information processing.

By measuring the extent to which left or right responses dominate relative to the total number of correct responses, the effect of the intercorrelation between the left and right measures should be eliminated. In table 2 correlations are small, but the tachistoscope circle index(TA-C-IN) seems to perform slightly better than the other, whereas the tachistoscope figure index (TA-F-IN) makes no sense at all. It even correlates negatively with the (DI-ML-IN). If any of these indexes were to be used in further work, it seems that among the tachistoscope indexes the circle (TA-C-IN) performs the best, whereas the two dichotic indexes do not differ. Based upon the evidence from table 1, however, and since the meaningless index is less influenced by uncontrolled left brain processing, this index would have to be preferred. However, the most striking observation to be made from table 2 are the very low correlations between the various measurements. And the same small correlations appear (0.17) when the combined tachistoscope lateralization score (VII) is related to the similar dichotic score (VII).

It is no surprise that researchers primarily basing their conclusions on tachistoscope measures often have disagreed with researchers basing their conclusions upon dichotic listening (see for instance Kimura, 1973). In the present study at least, it has not been possible to find evidence supporting that it is the same underlying phenomenon which is being measured with the two instruments.

TABLE 2

CORRELATIONS BETWEEN TACHISTOSCOPE AND DICHOTIC LATERALIZATION INDEXES

PSYCHOLOGICAL TESTS

In a previous paper (Hansen and Lundsgaard, 1981) 10 selected psychological tests were evaluated. A detailed description of these tests appear in this publication. Briefly, they include:

* 2 word mobilization tests, where the respondent is asked to name as many words he can come upon when: (1) the word street, and (2) animals are mentioned. Both tests are meant to measure left brain capability.

* 1 copy recall test measuring the ability to remember details of a story told to the respondent.

* 1 subtraction test measuring arithmetic capabilities.

* And finally, as the last measure of left brain capability: A word-pair learning test.

As measures of right brain capability were used:

* 1 face identification test, in which the respondent is supposed to point out pictures of the same person appearing on a piece of paper together with several other faces.

* 1 visual gestalts test measuring ability to recognize geometrical figures.

* 1 face recognition test measuring the ability to remember faces and rightly recognize them in a second presentation.

* 1 cube-test where the respondent is supposed to solve puzzles with cubes having various geometrical figures on their 6 sides.

* And finally, 1 incomplete picture test where the respondent is supposed to find missing items in two seemingly identical pictures.

TABLE 3

TWO FACTOR VARIMAX ROTATED PRINCIPAL COMPONENT SOLUTION FROM A STUDY OF LEFT AND RIGHT BRAIN CAPABILITIES OF 50 STUDENTS (Hansen and Lundsgaard, 1981)

In this study a factor analysis was conducted extracting two principle component factors which were varimax rotated. This solution is reproduced in table 3, where it appears that of the five left brain capability tests four of them correlate significantly with the first factor. Similarly four of the five right brain capacity tests correlate with the second factor. Two of the tests did not correlate strongly with either of the factors. Among the left brain tests, the subtraction test performs the weakest, and among the right brain tests, the incomplete picture test seems problematic since it loads strongly on factor one as well as on factor two.

In the present study the four first and the last four tests of table 3, were used. In addition to this, four more tests were included. The first digit span is borrowed from the Wechsler Adult Intelligence Scale. This is supposed to measure left brain capability by testing the respondent's ability to recall figures.

The other left brain capability test included is a simple arithmetic test using different combinations of addition and subtraction problems.

As a third new test it was decided to try the stroop test, since this test in a different context has been suggested as a measure of brain lateralization (Lindzay and Norman, 1977). The stroop test is interesting because it is measuring responses in situations where he or she is placed in a conflict between information submitted to the left and information submitted to the right hemisphere.

For example the respondent is asked to report what he sees, when the word green appears in front of him printed in a strong red color. Here left brain dominance should favor the response "green", whereas right brain dominance should favor the response "red".

Finally, as a supposed measure of right brain ability, a simple puzzle task was included.

All the 12 tests described here were administered to the same subjects, who had taken the tachistoscope and the dichotic listening tests. The resulting scores were standardized and normalized; and results from the eight tests included in the first study, also, were compared with the results from this study. No systematic differences in the right brain scores were found, whereas all left brain scores were higher for the university students than for the educational program students, two of them even significantly so. This is in accordance with expectations since the first group, is a group with more and presumably better school training, and, therefore, it should perform better on the "intelligence like" left brain tests.

TABLE 4

TWO FACTOR VARIMAX ROTATED PRINCIPAL COMPONENT SOLUTION OF 12 PSYCHOLOGICAL TESTS

The 12 psychological tests were factor analyzed. The two factor principal component varimax rotated solution is shown in table 4. Here, a somewhat mixed picture emerged. However, three of the four original left brain tests (copy, recall, and the 2 word mobilization tests) correlate with the first factor, and so does the stroop test. Similarly, two of the original right brain tests (visual gestalts and the cube-test) relate to the second factor, and so does the puzzle test. Problems appear, however, with the subtraction test (which was weak in the first test, also) and with the two new supposedly left brain measures, both of which correlate more with the second than with the first factor. Additionally the face identification test, which in the first application also correlated with the left brain factor, does so here again, and the same applies to the incomplete Picture test.

Following this a new factor analysis was carried out with the seven best tests out of which five are identical with those from the first application. For these tests a varimax rotated two factor solution was sought. This is shown in table 55 where it appears that the solution has some similarity with that of the previous application (table 3). Only the visual gestalts test does not fit this pattern. From table S it appears, that all tests loading strongly on the first factor, reflect left brain capabilities, whereas all tests loading on the second factor reflect right brain capabilities.

Altogether this gives some face validity to the interpretation of the two factors, as reflecting left and right brain capabilities. Similarly, the facts that the reduced test battery (table 5) produces factors comparing well with those of the first application (table 3) suggests that these tests have some reliability.

For further analyses a psychological test index was computed where the standardized and normalized scores on the 4 left brain items was deducted from the double scores on the 2 right brain items, and this again divided with the total score of all 6 tests.

TABLE 5

FACTOR LOADINGS FROM TWO FACTOR VARIMAX ROTATED PRINCIPAL COMPONENT SOLUTIONS BASED ON STANDARDIZED AND NORMALIZED RESPONSES TO 7 PSYCHOLOGICAL TESTS

THE RELATIONSHIP BETWEEN PSYCHOLOGICAL TESTS, DICHOTIC- AND TACHISTOSCOPE MEASURES

The psychological index does not correlate with either of the two most promising tachistoscope and dichotic tests. To examine the possibility of a relationship further, the seven psychological tests included in the last factor analysis (table 5) are correlated with the left and right visual field, measures from the tachistoscope circle and the meaningless word test (table 5). In interpreting these findings, it should be remembered that the ability to observe stimulation in the left visual field or by the left ear is supposed to indicate right brain processing and vice versa. In table 6 there are four strongly significant relationships (p < 0.01), and 3 relationships slightly less significant (p < 0.05). Most of these correlations, however, are opposing expectations, in the sense that they relate psychological tests measuring left brain capability with tachistoscope or dichotic measures supposed to indicate right brain capacity. To this observation should be added that already in the first study (Hansen and Lundsgaard, 1981) the relationship between dichotic listening measures and the psychological tests was questionable.

Therefore, it must be concluded that the psychological test do not validate the use of dichotic nor of the tachistoscope test as a measure for the identification of individual differences in lateralization in information processing. Of course it cannot be known whether it is the psychological test, the dichotic listening, the tachistoscope test or neither of them which actually measure such differences. We really only know that they do not measure the same phenomenon.

TABLE 6

CORRELATIONS BETWEEN SEVEN PSYCHOLOGICAL TESTS AND TACHISTOSCOPE AND DICHOTIC MEASURES

2. SELF-ADMINISTERED TESTS

The fourth type of measures included in the present study are self-administered tests. Here three different tests were used. From Richardson's (1977) verbalizer/visualizer test was included 15 of the original 20 items. To these, respondents should express agreement on a 5 point scale. The items are such as: "I like a job, where I am supposed to write and talk a lot (left)", "My imagination is better than of most others (right)". Of 15 items included, 8 were supposed to measure left brain capability.

It is worth-noting that the Richardson-test when it was developed, with some success, was compared with CLEM-measures. In later replications, however, this relationship did not hold up (Richardson, 1977).

Secondly, a forced choice test reported by Donegan (1979) was included. This test is composed of 20 items out of which 2 were deleted since they were assuming a knowledge of American culture which Danish subjects could not be expected to have. An example of the items which were used are

Which of the following two games would you prefer:

a) Scrabble (left brain)

b) Chesser (right brain)

If you were to solve a problem what would you prefer:

a) A cross word puzzle (left brain)

b) Picture puzzle (right brain)

Do you easily remember faces:

Yes (right brain)

No (left brain)

Do you easily remember names:

Yes (left brain)

No (right brain)

Finally, a test developed by the present authors was included. This test is composed of 40 statements rated on a S point Likert scale, ranging from agreeing to strongly disagreeing. The test is partly based on experiences from the study reported in Hansen and Lundsgaard (1981). The first 28 items are revised versions of selected items from this study. The remaining items are constructed as substitutes for those deleted from the first application.

Prior to its inclusion in the present study, the questionnaire was administered to a group of 103 evening graduate predominantly male business students, aged 25-40. Based o expectations regarding the effects of age, sex, and education, compared with the educational students, this group should be less right and more left brain oriented. Of the 24 significant differences found between the business graduates and the educational program students, 22 are in the expected direction.

For the Richardson and Donegan questionnaires a scoring plan is provided, so that for each of these a lateralization index can be computed. With regard to the Hansen and Lundsgaard questionnaire,several of the items are still tentative in character. Nevertheless, it is possible to hypothesize regarding their left vs. right brain relationship. Based upon this, it is possible to compute an left/right lateralization index for this questionnaire, also.

TABLE 7

CORRELATIONS BETWEEN RICHARDSON, DONEGAN, AND HANSEN/LUNDSGAARD INDEX

TABLE 8

CORRELATIONS BETWEEN 3 QUESTIONNAIRE INDEXES AND THE TACHISTOSCOPE AND DICHOTIC INDEXES

Correlations between these three indexes are shown in table 7. They are all significant and so large that it is reasonable to conclude that they all measure aspects of the same underlying psychological structure. When these indexes are compared with the tachistoscope and the dichotic listening scores, however, the very discouraging result of table 8 emerge. Only one correlation is significant and this is in a direction opposite to expectations. Based upon these findings it is hard to claim that any of the questionnaires measure the same phenomenon as is being measured with either the tachistoscope or the dichotic listening techniques.

TABLE 9

CORRELATIONS BETWEEN INDEX COMPUTED BASED UPON SELECTED PSYCHOLOGICAL TESTS AND SELF-ADMINISTERED TESTS.

Comparisons with the psychological test come out slightly more positively. The psychological index correlates significantly with two of the three questionnaire-indexes (table 9). With regard to the Hansen/Lundsgaard index, it should be remembered that 28 of the 40 items included in this questionnaire originally were selected based upon their relationship with the psychological test items in the first test reported in Hansen and Lundsgaard (1981). As pointed out earlier 4 items in the present psychological index are identical with those items appearing in the first test.

OTHER MEASURES

As discussed in the introduction conjugate lateral eye-movements (CLEM) have been applied in an attempt to measure tendencies for brain lateralization. In the Richardson (1977) application the relationship of the CLEM-measure to the self-administered questionnaire by Richardson is being studied, and even though the questionnaire is constructed based upon its relationship with the CLEM-measures, a revised relationship between the two measures is observed in later applications. Apart from this, we do not know, however, whether the CLEM-measure correlates with any of the measures tested in the present experiment.

With regard to the EEG-measures the situation is almost the same. In neither the Weinstein-studies nor in the Rockey, Green, and Perold-study are attempts made to validate the measurements with alternative measures of brain lateralization. The somewhat unclear results reporting in both studies, suggest problems with these measures, also. As discussed earlier, there are several methodological problems involved in connection with the EEG-measure. The extent to which the EEG-measure relates to either of the four measures used here, cannot be determined presently.

DISCUSSION

The present findings suggest that there are complicated problems involved in studying brain lateralization in an economic psychological context. It is warranted to recommend that much more effort is put into the development of suitable measurement techniques before theoretical construct derived from brain lateralization research can be of any use in the study of economic psychological behavior. They also suggest that findings reported by researchers in different research traditions should be viewed with caution. The results reached in other areas than economic psychology may be as dependent upon the measurement techniques as the findings reported here suggest. One must be cautious with regard to the interpretation of the relationship between the actual results and the underlying brain half specialization.

Also, the present findings make it important to emphasize the need for detailed methodological information to be included in all reports. This of course should apply to all good research. It is, however, not always being followed, and in the present area this is highly, unfortunate. Among the very large number of papers being reported in the psychological literature where dichotic listening has been applied for the study of brain lateralization, very few can be found where thorough replication can be carried out based upon the methodological information presented together with the findings alone.

The methodological problems are so much more important since the techniques used in those studies from where most of our knowledge about brain lateralization comes, are not easily - if at all - applicable on larger samples or even in smaller laboratory studies in economic psychology. This applies to split-brain observations, studies of brain damages, the use of electro-shock, and brain blood-flow photography.

Therefore, it is necessary to be sure that those substitute techniques, which are applied, do measure the same phenomenon which is being studied in more basic brain lateralization research.

In the view of the present authors, the discouraging findings reported here should not be taken as to proof that brain lateralization is an uninteresting phenomenon to the economic psychologists. On the contrary, brain lateralization research has a potential for making it possible to approach problems which have so far been unattainable to the economic psychologist. Therefore, it is of the greatest importance to solve those methodological problems which have been pointed out in this paper. To do so, however, requires that efforts are concentrated on this, rather than wasted on speculative applications, resulting in findings, the interpretation of which are far from obvious.

REFERENCES

Appel, Valentine, Weinstein, Sidney, and Weinstein, Curt (1979), "Brain Activity and Recall of TV-advertising," Journal of Advertising Research, 19, pp 7-15.

Bogen, Joseph E (1977), "Some Educational Implications of Hemispheric Specialization," (ed.)" C Wittrock, in The Human Brain (New York) pp. 133-152.

Deglin, Vadim L (January 1976), "Journey Through The Brain," (Unesco Courier) pp 4-14.

Donegan, Frank (November 1979), "Split Right Down the Middle-Which Side of Your Brain Is Working?" Psychographic Self-Examination Series (Penthouse) pp 95-98.

Hansen, Flemming (June 1981), "Hemispheral Lateralization Implications for Understanding Consumer Behavior," The Journal of Consumer Research, pp. 23-37.

Hansen, Flemming and Lundsgaard, Niels Erik (April 1981), "Brain Lateralization and Individual Differences in People's Reaction to Mass Communication," ESOMAR Conference.

Kiruma, Doreen (1973), "The Asymmetry of the Human Brain-Recent Progress in Perception," Readings from Scientific American (San Francisco).

Kling, Nathan (1980), "Hemispheric Asymmetries Relative Cerebral Hemispheric Dominance and Information Usage," (paper presented at Atlantic Schools of Business Conference, Halifax).

Krugman, Herbert E. (February 980), "Substained Viewing of Television," Paper presented at the Conference Board, Council on Marketing Research (New York)

Lindzay, Peter H., and Norman, Donald A (1977), "Human Information Processing," (Nev York Academic Press)

Richardson, Alan (1977), "Verbalizer-Visualizer A Cognitive Style Dimension," Journal of Mental Imagery, 15 pp 109-26.

Rockey, E. A,. Greene, W F., and Periold, E. A (1980), "Attention, Memory and Attitudinal Reactions to Television Commercials Under Single and Multiple Exposures Conditions as Measures by Brain Research," (paper presented st the ACR 26th Annual Conference).

Sperry, Roger E. (1973), "Lateral Specialization of Cerebral Function in the Surgically Separated Hemispheres," (eds.) F J. McGuian, and R. A. Schoonorer, in The Psychophysiology of Thinking (New York: Academic Press).

Weinstein, Sidney, Weinstein, Curt, and Appel, Valentine (June 1980), "Brain Activity Responses to Magazine and Television Advertising," Journal of Advertising Research, 20.

Wexler, Bruce E.(1980), "Cerebral Laterality and Psychiatry: A Review of the Literature," The American Journal of Psychiatry, 137:3, pp 279-289.

Wexler, Bruce E. and Heninger, George R. (March 1979), "Alterations in Cerebral Laterality During Acute Psycho tic Illness," Arc Gen Psychiatry, Vol 36, pp 278-284.

Wittrock, Merlin C (1977), The Human Brain (Prentice Hall: Englewood Cliffs, NY).

----------------------------------------