Assignment #1 DESCRIPTIVE STATISTICS & COMPARISON OF MEANS

Dr. Z Fall 2018

SSCI 3910

FACULTY OF SOCIAL SCIENCE & HUMANITIES FALL 2018

OPTIONAL ASSIGNMENT #1

Assignment #1 DESCRIPTIVE STATISTICS & COMPARISON OF MEANS TOTAL: 20% of Final Grade

DUE DATE: WEEK OF NOVEMBER 5, 2018 AT THE BEGINNING OF LAB

NO MORE THAN 8 PAGES OF TEXT (NOT INCLUDING TITLE PAGE/REFERENCES) Complete this assignment using ‘DEMO1.SAV’ secondary data set using the variables below.

RACE OF RESPONDENT (RACE) HIGHEST YEAR OF EDUCATION (EDUC) HOMOSEXUAL RELATIONS (HOMOSEX)

1. Identify the levels of measurement and the type of variables you are to analyze, as well as what your ‘factors’ and dependent variable are. 2. In SPSS, run the most appropriate univariate statistics for RACE, EDUC & HOMOSEXUAL RELATIONS in accordance with levels of measurement. Remember, to think about any recoding or data modification you may have to do (or not). Provide univariates for both original and recoded variables. 3. In SPSS, run a 2x2 zero-order crosstabulation with at least two of the above variables to test the statistical dependency of variables. This relationship must statistically significant. Ensure that the relationship you are testing makes conceptual sense. 4. In SPSS, run a 2-Way Factorial ANOVA using RACE, EDUC & HOMOSEX. Be sure to include the post hoc tests (Bonferroni). Include any POSTHOC test. 5. You need to interpret the entire analysis. Interpret your findings using headings and subheadings and write up your analysis as per course outline. Provide both technical and substantive interpretations. Be sure to discuss ALL the appropriate statistics. -Make sure you include hypotheses -Discuss all relevant statistics associated with the test(s) -Use subheadings to present your assignment.

Title Page: A proper title that speaks to your analysis and variables. Research Problem: Introduction of topic area, statement of research question(s) to be addressed (including any sub- questions with controls); Statement of the independent and dependent variables to be investigated, statement of control variables (if applicable). For each variable: the question asked in the survey, the response categories and the value labels for each response category, level of measurement (i.e.: nominal, ordinal, I/R) and type of variable (i.e.: string, categorical, discrete or continuous) with explanation and justification for responses with respect to criteria for the level of measurement and type of variable.

Dr. Z Fall 2018

Hypotheses (only if a bi/multivariate model is being tested): Outline of specific hypotheses being tested (i.e.: null and research hypotheses) as is appropriate for EACH statistical procedure used; inclusion of a causal (arrow) diagram modeling the relationship(s) being tested, statement of causal explanation or justification of why and how the variables might be related, if the test is based on cause and effect, like crosstabs and regression. Method: Secondary data analysis explained (i.e.: what is secondary data and why is it being used ex: existing data collected by someone else and they are used for convenience and low cost), statement of data set that will be used to investigate the hypotheses, why the data set was chosen (ex: because it included the specific variables related to my research topic/research question). Data Modification(s), if any: Explanation of need to recode/scale any of the variables (ex: why recoding was necessary, how new variable and categories were created from old variable, and value labels assigned to new categories, level of measurement and type of variable for all recoded variables). This should be done as a table. Results: Results of statistical analysis are DISCUSSED AND INTERPRETED. Both technical (i.e.: numerical) results are reported and substantive messages (i.e.: interpretation in words of what the numbers imply/mean/represent) are articulated Conclusion: Technical and substantive summary of ONLY key findings (i.e.: both significant and non-significant relationships found and the numeric results/key statistics which evidence/indicate these findings). Relates these findings back to the null hypotheses for EACH statistical procedure. Relates the findings back to the main research question and sub-questions posed. No discussion of numbers here please. SPSS Output and Formatting: Output that is CORRECT and COMPLETE including original and recoded variables, definition and labelling of recoded variables with appropriate titles; Graphs titled and numbered; typed, d/spaced, stapled securely with page numbers and 1” margins, subheadings and references.