Intimate Partner Violence: Another Look

  • Due May 9 by 11:59pm
  • Points 5
  • Available until May 9 at 11:59pm

Intersectionality:  Gender and Race/Ethnicity

In Download Conditional Probability - Violence against Women.docx

we saw that the probability that a woman reported violence at the hands of an intimate partner was different, depending on their "race"/ethnicity.   Here is the table:   

Intimate Partner Violence and Race/Ethnicity
Whether experienced IPB Black/African-American American Indian/Alaska Native (Indigenous) Hispanic White Other ROW TOTALS
Experienced IPV 55 4 49 239 19 366
Not experienced IPV 69 4 83 444 34 634
COLUMN TOTALS 124 8 132 683 53 1000

Source:  https://everytownresearch.org/reports/guns-intimate-partner-violence

Question 1:  These results were based on a SAMPLE of 1000 women in 2019.  If a different sample of 1000 women had been randomly selected, would you expect the numbers to be the same?  Why or why not?

Inferences from a Sample to a Population

Due to sampling error, we would NOT expect the same numbers from every random sample of 1000 women from 2019.  But would the same patterns hold for the POPULATION of all women in 2019?  That is a HYPOTHESIS TEST question.

The appropriate test is a "Chi-Squared Test of Independence."    The null hypothesis is that the variables "race/ethnicity" and "experienced intimate partner violence" are INDEPENDENT.  The alternative hypothesis is that these variables are NOT INDEPENDENT (that is, that there is a relationship),   

To conduct this test, the "expected values" in every cell of the table need to be at least 5.  To fulfill this condition, we will have to combine the Indigenous women with "Other" (which also includes Asian/Pacific Islander women).

The revised table is:

Intimate Partner Violence and Race/Ethnicity - Revised
Whether experienced IPV Black/African-American Hispanic White Other (including Indigenous) ROW TOTALS
Experienced IPV 55 49 239 23 366
Not experienced IPV 69 83 444 38 634
COLUMN TOTALS 124 132 683 61 1000

Question 2:  Conduct a Chi-Squared Test of Independence to decide, at a 5% level of significance, whether or not the variables "race/ethnicity" and "experienced IPV" are independent in the population from which this sample was drawn.  Make sure to include all steps of the Hypothesis Test, including a clear statement of the result.

Post comments, answers to questions, etc. in the Intimate Partner Violence: Another Look Discussion

 

1746860399 05/09/2025 11:59pm