H. Wayne Horgan (1973) carried out eleven separate studies that involved 881 males and 1021 females, of which all were of white origin. When asked the question “ What do you estimate your IQ score would be should you take one of the standard objective tests?” it was found that females underestimated their own IQ. Five of the eleven studies carried out were significant at the five percent level. When the participants were then asked to estimate their parents IQ, the females as well as the males gave a higher IQ estimation for their fathers.
Louise Higgins (1987) study supported these findings. When she carried out a study at Chester College in England, women’s mean IQ estimate was 111 compared to the men’s mean IQ estimate of 118.
Tomi-Ann Roberts (1991) study at Stanford University explained such findings. She stated that men tend to respond to only positive feedback, whereas women respond to both negative and positive feedback. This is because men have a more competitive nature and so they tend to ignore information that they find dissatisfying. This in turn means that women are more reasonable in thinking about logical solutions, whereas men are more rationalising.
The study being carried out will be a replication of Horgan (1973), Higgins (1987) and Belloff (1987 – 1991) because they all initiated the idea that females estimated their own IQ lower than that of males.
The aim of the study is to find out if males score higher than females when asked to estimate their IQ.
Males will estimate their IQ higher than that of females. This is a one-tailed hypothesis since the direction is given.
METHOD
Design
The design of the experiment used will be an independent measures design. This is because the two groups (males and females) in this experiment consist of different individuals. Two groups were used in this experiment. One group was male, and one was female. Participants were allocated into the two groups depending on their gender. The independent variable (IV) in this experiment is gender. The dependent variable (DV) is the estimation of IQ. The controls used in this experiment were the standardised procedures and instructions (see appendix 1, page 12).
Participants
Participants were individuals in higher education aged 18-25 years old. They were chosen using opportunity sampling. The total number of participants used was 20, 10 of which were male and 10 of which were females.
Materials/Apparatus
A set of standard procedures and instructions were used (see appendix 1, page 12). A table was produced in order to record the findings of all 20 participants (see appendix 2, page 13).
Procedure
Four experimenters were used. Each of which had to get 5 participants each. If participants had already taken an IQ test and know their IQ then they could not be used in the experiment. Each participant was asked the standardised questions and debriefed at the end (see appendix 1, page 12). Results were recorded in a table for later comparison (see appendix 2, page 13).
RESULTS
The data was put into two tables, one for female, and one for male (see appendix 3, page 14). The measures of central tendency used to calculate the data were the mean. This allows the average of the values to be determined for later comparison. The measures of dispersion used were the range and standard deviation. This allows the spread of the values to be seen and compared.
The value of the mean was calculated using a formula. The means for females was 102.5 and for males were 111.2 (see appendix 4, page 15 for calculations).
The range was calculated using a formula. The ranges for females were 46 and for males were 55. The standard deviation formula was then used to calculate the measure of spread. The standard deviation for females was 15.71 and for males was 17.98 (see appendix 5, page 16 for calculations).
Non-parametric tests of difference was used therefore data had to be ranked from lowest to highest (see appendix 6, page 17). Since an independent measures design was used, a Mann-Whitney ‘U’ test was used. The value of U was calculated (see appendix 7, page 18) which was 35.5. This value of U was compared against the value in the Mann-Whitney table at the 0.05 percent significant level (see Coolican, H. 1999. Research Methods and Statistics in Psychology: 3rd edition. Page 556). The value stated in the table is 27. Since the value of U from this experiment was calculated to be 35.5, this means that it is not significant as it is higher. This means that the hypothesis has to be rejected since males did not estimate their IQ higher than females.
DISSCUSSION
The aim of this study was to find out if males estimate their own IQ higher than females. Since the value of U, which was 35.5, was higher than the value stated in the table, which was 27, it can be said that the results found were not significant at the 0.05 percent significant level. This means that the hypothesis has to be rejected since males did not estimate their IQ higher than females. The findings therefore do not support Horgan (1973), Higgins (1987) and Belloff (1987 – 1991) because they all found that females estimated their own IQ lower than that of males.
There are many reasons as to why the study does not support Horgan (1973), Higgins (1987) and Belloff (1987 – 1991) findings. For example, the studies carried out previously were over twenty years ago. Women these days are more confident and independent so this can have an effect. Also, equal rights can have an effect on the level of confidence a woman has.
There were many problems with this study that can account for the findings. Since there were four experimenters, the way in which the standardised questions asked could have still differed. For example, tone of voice could have an influence. Time of day can also have an effect since participants are less likely to give their best towards the end of the day because they may be tired or get bored easily, therefore the ‘screw you’ effect would occur. Limitations of age could have affected the results because it age range used was 18-25 years old. Academic confidence would play an important factor as well because it is more likely that the older the participant was, the higher they would perceive their IQ. Low self-esteem can result to underestimations of IQ.
The type of design used in this study had disadvantages. This was because there was a wide range of variability with participants. For example, age, academic confidence, self-esteem. This could be avoided by using random allocation.
Using opportunity sampling had major disadvantages. This is because the participants used were all from one area and so this cannot give a true representation of the population as a whole.
If this experiment was to be carried out again, many of the problems stated could be avoided. One experimenter would be used so that the way in which the standardised questions are used will be the same. The time of day in which the study was carried out can be kept as another control. The environment in which the study was carried out should also be kept constant. For example, somewhere where disturbances are less likely to occur. A larger sample of about 50 to 100 would provide a better representation of the population. Random sampling should be used instead of opportunity sampling so that it gives a better representation of the population. Using computer selections, random number tables and manual selection could carry this out. More questions should be asked as a basis. For example the participants could be asked if they actually knew what IQ meant and their perception of it.
Further research that could be carried out could involve testing different factors that can affect IQ estimation. This could be race, culture and even age. Since underestimation of IQ tests is a common problem as found by Zingler et al (1973). It was that found pre-school children from poor backgrounds showed at least a 10-point gain in second IQ tests or after a play session. Where as, middle-class children achieved a 3-point gain. Therefore, it can be argued that even actual testing of IQ can have an effect on the estimation of a persons IQ.
CONCLUSION
The hypothesis has to be rejected since males did not estimate their IQ higher than females. The findings therefore do not support Horgan (1973), Higgins (1987) and Belloff (1987 – 1991). This can be due to the fact that equal opportunities, confidence and independence in women can account for this. Further research can be carried out using race, culture, and age with the estimation of IQ.
REFERENCES
Bellof, H. (1992). Mother, Father and Me: Our IQ. The Psychologist, 5, 309-311.
Coolican, H. (1999). Research Methods and Statisitcs in Psychology (3rd edition). London: Hodder and Stoughton.
Eysenck, M. (2002). Psychology: A Students Handbook. Psychology Press Ltd.
Higgins, L. (1987). The Unknowing of Intelligence. The Guardian. 10th February 1987.
Hogan, H. W. (1978). IQ Self Estimates of Males and Females. Journal of Social Psychology, 106, 137-138
APPENDIX 1
I am a student conducting a study on the perception of IQ. Would you mind if I asked you a few questions?
- Have you ever studied psychology before?
- Do you know your own IQ score?
- How old are you?
- The national average IQ is 100. Based upon this, how would you estimate your own in numerical form?
Debrief:
The purpose of the study was to distinguish how IQ relates to gender. Would you still want your details to be used, as we will be analysing them? Thank you for your support.
APPENDIX 2
APPENDIX 3
APEENDIX 4
APPENDIX 5
APPENDIX 6
APPENDIX 7