DV is a two-category variable and we can assume there are three IVs that are categorical. To see the effect of IVs on DV here Logistic Regression Analysis can be applied. Kindly read the chapter in any stats book.
If your dependent variables are 'interval' (continuous) measures, run factorial ANOVAs with 3 independent 'factors': Age Group, Gender and Ethnicity. If the 2 dependent variables (punitive vs. rehabilitation attitudes) are on the same scale, you can test the differences between both with repeated measures ANOVAs. Your 'final' design would then be a Within (2-levels factor: punishment vs. rehabititation) x Between (Age group) x Between (Gender) x Between (Ethnicity) subjects design. Be sure to check the 'normality' of the distributions and equality of variances of your dependent variables with histograms, Shapiro and Levene tests.
h1) Stronger punitive attitudes and weaker rehabilitation attitudes toward sex offenders. h2) Stronger rehabilitation attitudes and weaker punitive attitudes toward non-sexual offenders. h3) a) Older participants will have stronger punitive attitudes than younger participants. b) older participants will have weaker rehabilitation attitudes than younger participants. h4) Females will have stronger punitive attitudes than males, and weaker rehabilitation attitudes than males. h5) Individuals who identify as white will have stronger punitive attitudes than other ethnicities and weaker rehabilitation attitudes.
The IV's are gender (male, female), age (18-25, 26-35, 36-45, 45&older), and Ethnicity (Caucasian/White, African British/Black, Asian British/Asian, Arab British/Arab, and others)
The DV's are Punishment attitudes (4 items questionnaire, likert scale 1-strongly disagree to 4- strongly agree).
And rehabilitation attitudes (4 items questionnaire, likert scale 1-strongly disagree to 4- strongly agree).