# [[EDUC 7215]] Assignment 10 Jethro Jones Also available online at [drjethro.com](https://drjethro.com/7215ass10) # Assignment ### Introduction - filter education groups 2,3,4,5 (2=HS grad or GED, 3=Associate degree or certificate/license, 4=Four-year degree, 5=Masters, doctoral or professional degree) - Filter gender 0=males, and 1=females. - (education=2 or education=5) and gender ne 3. - two-way frequency table - Chi-square tests - Fisher’s Exact test p-values - Calculate and interpret relative risk to support your conclusion ### Research Question Is there a relationship between gender (male & female) and education (HS or GED & Masters/PhD/Professional Degree) for survey respondents in the Missouri area? ### Hypotheses H₀: Education is independent of Gender (no association between education and gender), H_a: Education is dependent on Gender (an association exists between education and gender). ### Type 1 Error Rate 0.05 or 5% ### Methods I used Tasks and Utilities -> Tasks -> Statistics -> Table Analysis to generate the two-way frequency tables and their statistics. I selected the genderEd dataset from our library. Under Roles, I selected study and gender as the row variables, and graduate as the column variable. Under the Options tab, I checked the boxes for Suppress Plots, Observed Frequencies, Row & Column Percentages. I also checked the boxes for Chi-square Statistics and Odds Ratio & Relative Risk. After I clicked the Run icon, I opened it in a new window and took screenshots, then used an AI tool to rename those screenshots so they would show up here in this report. I also generated custom formats with the SAS Code below: ``` /* Step 1: Filter data for education = 2 (HS or GED) or 5 (Masters/PhD/Professional) and gender = 0 (male) or 1 (female), excluding non-binary (3) */ DATA FilteredData; SET Behavior; IF (education=2 OR education=5) AND gender NE 3; RUN; /* Step 2: Frequency analysis to explore the relationship between gender and education */ PROC FREQ DATA=FilteredData; TABLES gender*education / CHISQ NOROW NOCOL NOPERCENT; TITLE "Relationship Between Gender and Education Level (Filtered Data)"; RUN; ``` The tables generated from both the code and the GUI resulted in identical graphs. ### Summary Statistics The study had about 25% more females than males, with n=126 females and n=100 males. 62% of males had higher education and 75.4% of females had higher education. When comparing across gender, 60.51% of those with higher education degrees were female, while 39.49% were male. ![[7215 ass10 2025-04-21 GenderByEducationTable.png]] ### Hypothesis test for higher education and gender I analyzed if education was related to gender. the two-way frequency table seems to indicate that higher education attainment is lower for men than women. More women have degrees than men, and a higher percentage of women surveyed received degrees than men surveyed. H₀: Education is independent of Gender (no association between education and gender), H_a: Education is dependent on Gender (an association exists between education and gender). ⍺ = Type I error = 0.05. Decision Rule: If p-value < 0.05 then reject H₀, otherwise, do not reject H₀. Conclusion: since p=0.0299, p=0.0301, p=0.0427, and p=0.0302 on the Chi-Square tests**All four p-values are < 0.05**, so we **reject the null hypothesis**. There **is a statistically significant association** between **gender and education level** in this dataset. ![[7215 ass10 2025-04-21 Statistics Table Gender Education.png]] ### Fisher’s Exact Turning to the Fishers exact test, we see that the two-sided Pr =.0414 so this test also concludes that Education is dependent on Gender, and we can reject the H₀. ![[7215 ass10 2025-04-21 FishersExactTestResults.png]] ### Relative Risk table Odds Ratio = 1.8783 (95% CI: 1.0599 – 3.3285) The odds of a female having a post-baccalaureate degree are 1.88 times the odds for a male. Since the confidence interval does not include 1, this is statistically significant. Relative Risk (female vs. male) = 1.5445 (CI: 1.0404 – 2.2929) Females are 54% more likely than males to have a post-baccalaureate degree. ![[7215 ass10 2025-04-21 Odds Ratio and Relative Risks Table.png]] ### Conclusion These tests confirm that in the Missouri area, women are significantly more likely than men to hold advanced degrees in this sample. There is statistically significant evidence that gender and education level are related in this dataset. Specifically, females are more likely than males to have post-baccalaureate degrees, as supported by the two-way table, Chi-square test, Fisher’s test, and odds ratios. ### SAS Screens: ![[CleanShot 2025-04-21 at [email protected]]] ![[CleanShot 2025-04-21 at [email protected]]] ![[CleanShot 2025-04-21 at [email protected]]]