EDP 362T Lecture Notes - Lecture 4: Null Hypothesis, Liberal Feminism, Mean Absolute Difference
Lecture 4:
The Politics of Difference and Similarity:
• Opposite Sex
• Similarities tradition – claim that women and men are very much alike in
intelligence, personality, abilities, and goals. Stems from liberal feminism, argue
for equality for all sexes.
• Difference tradition – claim that there are fundamental differences between
women and men that should be recognized and honored. Stems from cultural
feminism, taking care of other people and relationships was rewarded as
dominance and personal ambition.
o Both are used for research and political strategies = richer and more
flexible.
o Has controversies in defining and measuring difference, values and
interpretations of results.
Defining Difference:
• No one has ever discovered a psychological trait or cognitive ability on which
men and women are completely different.
• Average or mean difference – doesn’t tell much info by itself. Set of distributions
can have same difference in mean but large variability, and the meaning of
difference is different for each distribution.
• Statistical significance – not the same as importance; unlikely for the differences
to have occurred by chance.
• Practical significance – answers more general questions about whether
difference is large enough to be of value in a practical sense.
Measuring Differences:
• Mean difference – an average difference.
• Variability – range/spread of data.
• Statistically significant – unlikely to have occurred by chance.
• Confounding – the effects of 2+ variables are mixed, and it becomes impossible
to determine which variable is causing experimental effects.
• Meta-analysis – using quantitative methods to summarize the results of research
conducted by different people at different times. Assess magnitude and
consistency of different effects statistically.
o Helps interpret data from large numbers of studies and estimate
difference; simplifies study of other variables interacting with gender.
o Can’t compensate for biases in original studies, nor ensure objective
interpretations. Researchers themselves must decide which are relevant.
o Can show which variables moderate the occurrence of gender differences,
but can’t determine the causes for difference.
• Moderator variable – variable that interacts with another variable to change its
effect.
• Logic of experimental design contributes to emphasis on difference instead of
similarities. When found diff = confident it exists, more likely to be published =
overemphasis leads to limitation of hypothesis testing.
o Research hypothesis: HA =! 0
o Null hypothesis: Ho = 0
find more resources at oneclass.com
find more resources at oneclass.com
Gender and Math Abilities:
• When trying to disprove theories of inferiority of women to men, research is set
up to find difference.
• A lot of difference focuses on biology, which is harder to change. Have people in
powerful positions who don’t understand or haven’t done research perpetuate
these differences.
• In the past, women have been deemed inferior by the dominant group to
maintain status quo. Even done so by scientists.
o Female brain: different and inferior. Tried to document by race in terms of
brain size (brain-size hypothesis: black<Asian<white). Measured
intellectual capacity by measuring ratio of brain size to body but didn’t
work.
o Male brain: variability hypothesis = men are more variable, allows for more
advantageous characteristics for evolution, explains why there are more
intelligent men than women.
• General pattern of cognitive skills is one of gender similarity.
• Boys score higher on standardized tests and girls are higher in school
achievement. Girls not viewed as intelligent for this.
o When boys fail, treated as unfortunate and unlucky.
o When girls fail, treated as not smart enough.
Factors influencing math performance:
• Girls perform better on standardized math tests in elementary school but trend is
reversed in high school.
• Biological perspectives: genetic contributions of gender, hormones, and brain
structure.
• Math as a male domain: stereotypes, if females go into this field, seen as
unfeminine and cold. Men holds beliefs of this stereotype stronger than females –
others’ beliefs can create self-fulfilling prophecies.
• Learning the lesson: I’m just not that good at math.
o Classroom: benevolent sexism to girls decreases girls’ ability to do well
(not picking on girls because don’t want them to be embarrassed when
getting answer wrong) + hostile sexism (you don’t belong in this
classroom, should be home training to be housewives; threatened by the
competence of women).
o Parents have an impact on your beliefs/expectation on yourself: when girls
are struggling, would intervene and help early, not develop analysis skills.
When boys are struggling, let them tough it out.
• Stereotype threat: contribute to gender gap, when girls believe in the stereotype,
do worse, but if believe they’ll do better than men, will do better. Can cause
stress-related physiological responses, requiring actively suppressing negative
thoughts/emotions which disrupt working memory and interfere with problem-
solving ability. Happens more for women in male-dominated areas of study.
• Hyde found negligible difference for gender difference in outlying math scores,
but boys slightly outnumber girls in the 99th percentile.
o Equilizing scores: girls are taking more advanced math courses, and
women receive nearly half of all bachelor degrees given in math in the US.
find more resources at oneclass.com
find more resources at oneclass.com
Document Summary
The politics of difference and similarity: opposite sex, similarities tradition claim that women and men are very much alike in intelligence, personality, abilities, and goals. Stems from liberal feminism, argue for equality for all sexes: difference tradition claim that there are fundamental differences between women and men that should be recognized and honored. Defining difference: no one has ever discovered a psychological trait or cognitive ability on which men and women are completely different, average or mean difference doesn"t tell much info by itself. Assess magnitude and consistency of different effects statistically: helps interpret data from large numbers of studies and estimate difference; simplifies study of other variables interacting with gender, can"t compensate for biases in original studies, nor ensure objective interpretations. When found diff = confident it exists, more likely to be published = overemphasis leads to limitation of hypothesis testing: research hypothesis: ha =!
