In: Statistics and Probability
Your Task
Below are some fictitious data consistent with the findings of Nisbett and Wilson. For the sake of
simplicity, we’ll only analyze the data for “Physical Appearance.” Physical appearance was measured
using five traits with each trait rated on a 5‐point scale (larger numbers indicate higher ratings of
physical attractiveness). Participants’ answers were added together to arrive at a final “physical
attractiveness” score that could range from 5 to 25. These data are presented in table 1.
Your task is to statistically analyze these data and decide whether or not there is evidence here for the
“halo effect.” A series of questions have been provided to help guide your exploration of the data.
Table 1. Ratings of Physical Attractiveness
Type of Exposure
‘Warm and Friendly Instructor’ ‘Cold and Distant Instructor’
23 17
17 12
20 9
18 13
23 10
20 15
19 14
24 11
17 15
20 13
Short Answers (total 400 – 500 words)
Type your answers. You should include the question number for each answer, but you do not need to
include the question itself.
1. In your own words, describe the halo effect and provide a specific example.
2. In this study, what was the independent and dependent variable?
3. What would be the null hypothesis for this analysis? Why do we assume the null hypothesis is
true when doing inferential statistics?
4. What is the alternative (or research) hypothesis?
5. Describe the center of each distribution: What is the mean for each group?
6. Describe the variability in each distribution: What is the standard deviation for each group?
What does this measure tell us about the dispersion of scores in each group?
7. These data should be analyzed using a t‐test, but which one should be used (i.e., one sample t
test, t test for independent samples, or t test for related samples) and why?
8. Should you conduct a one‐ or two‐tailed test? Explain your answer.
9. Describe the results of the t test. Were the results statistically significant at p < .05? Explain
how you arrived at this decision (i.e., what was the t value you calculated)
10. From all this information, can you conclude that the “halo effect” was present in this study?
Was the researchers’ hypothesis supported? Explain your reasoning.
Reference
Nisbett, R. E., & Wilson, T. D. (1977). The halo effect: Evidence for unconscious alteration of judgments.
Journal of
1. Halo effect:
"The idea that global evaluations about a person (for examples, she is likable) bleed over into judgments about their specific traits (for example, she is intelligent)"
2.
Independent variable = Whether students saw a video of an instructor whose manner was warm and friendly or cold and distant.
Dependent variable = Ratings of physical appearance.
3. Null hypothesis: There is no difference in physical appearance ratings between the two conditions.
4. Alternative hypothesis: Physical appearance ratings will be higher in the warm than in the cold condition.
5. Center of each distribution is mean.
The formula of a mean is:
here n = 10
Mean for warm:
Mean for cold:
6. The variability of distribution will be measured using standard deviation.
The formula of standard deviation is,
Subtract the mean from each observation then take a square and then take a sum of that all values to get the numerator
SD for warm:
SD for cold:
The variability of scores in each group is about the same which tells us that the scores in each group cluster in similar ways around the group mean.
7. t-test for the independent sample would be used since both the data sets are independent.
8. This is a one-tailed test since here have to check the halo effect on weather conditions that are physical appearance will be more in the warm than cold.
9. The formula of t-test statistics is
Where
t-test statistics = 6.47
P-value = P(t > test statistics) = P(t > 6.47)
By using excel P- value for test statistics 6.47 and degrees of freedom = n1 + n2 - 2 = 18 is, 0.000002 that is approximately 0
So here P-value < 0.05, so the test is statistically significant.
That is rejecting the null hypothesis at 0.05 level of significance because P-value is less than alpha(0.05)
10. Yes, the result was consistent with the halo effect, that is the halo effect was present in the study. The general positive evaluation associated with interpersonal warmth (as opposed to the general negative evaluation associated with coldness) bled over into higher ratings of physical attractiveness.