Research Paper Using SPSS 4500 WORDS

2

Final Paper Checklist

· Intro

· Grabs attention.

· Makes subject relevant to today (i.e., Why should we care?)

· Clearly states what relationship you are investigating (e.g., This research will investigate the relationship between [state the independent variable] and [state the dependent variable]).

· Lit Review Analysis – for each empirical piece (four minimum) discuss:

· Purpose of the study?

· Who conducted the study?

· When was the study conducted?

· Where did the data come from?

· How was the data sampled?

· When was the data collected?

· How was the data collected?

· Size of the sample?

· Margin of error for the sample (for surveys only)?

· Key variables (ones that equate to your independent and dependent variables) and how they were conceptualized/defined and operationalized/measured?

· List of all other independent variables?

· What did the researcher find with emphasis on whether the relationship between your variables of interest was statistically significant and, if so, the measure of association/strength of the relationship (e.g., too weak to be of any importance, weak, moderate, strong)? Try to provide statistical test evidence for the findings such as r2=.42, p=.000. Any other significance findings that are relevant to your study?

· Provide a VERY succinct description of researchers’ argument as to why the relationship exists or doesn’t exist between your independent and dependent variable, based on the evidence.

· Discuss the piece’s shortcomings, if any, you feel need to be noted about the findings or any other aspect of the research design (note: this is particularly important when you include a finding that is contrary to the expected outcome or other outcomes you are including as evidence). Remember: unless you are specifically noting shortcomings to discount a piece’s findings that are contrary to your hypothesis, you shouldn’t have very many shortcomings or you need to rethink whether it’s a strong enough piece to be included in your study! If your pieces are strong, there may not be any shortcomings to note.

· Lit Review Conclusion

· Summarize and compare and contrast all unique findings from your lit review. Make sure to cite appropriately and often.

· Bridge to Methods and Findings Sections

· Add your purpose (e.g., “The next section of this paper specifically tested whether there is a relationship between religiosity and support for same sex marriage.”

· Data Source(s) Discussion: Discuss data source(s) before univariate analysis if there is one source for all variables such as when the individual is the unit of analysis and you are using survey data such as the GSS. If you have a different data source for each variable, discuss at the beginning of your univariate analysis for each variable.

· What is the data source?

· What is the year(s) of the data?

· Who or what org collected it?

· How long has this type of data been collected?

· How was it collected?

· How was it sampled?

· Size of sample?

· Margin of error for the sample if the source is a survey?

· Any other info you feel helps make a case for reliability and validity?

· Univariate Analysis (minimum of two - one for your dependent variable and one for your independent variable). If you include a control variable or additional independent variables, you must conduct univariate analysis for ALL variables included in your study. Complete univariate analysis in whole for each variable – don’t jump back and forth between variables with each piece of evidence.

· Conceptualize your variable (i.e., define the concept – don’t use a tautology and don’t confuse this with how the variable was operationalized/measured; use a conceptualization from one of the pieces in you lit review or a dictionary and paraphrase and cite.

· Operationalize your variable including providing the available categories if the variable is nominal or ordinal.

· Provide the one best central tendency measure (CTM) with appropriate statistical test evidence (e.g., mode, median, or mean). If you have ratio/interval variables with an extreme outlier(s), we need to talk before you continue with your analysis!

· Analyze dispersion/variability with appropriate evidence (e.g., description of frequency distribution for nominal and ordinal variables using frequency percentages to describe as low, moderate, or high variability; standard deviation for ratio/interval variables without extreme outliers).

· Calculate the 68% data range for Ratio/Interval level variables only (i.e., add and subtract one standard deviation from the mean to calculate the range where 68% of the cases lie).

· Insert a frequency table with a reference in your text (make sure data is aligned under headers in your table).

· Insert an appropriate graphic with a reference in your text (you may not be able to get a graphic for ratio/interval level variables).

· Bivariate Analysis

· Provide your hypothesis (H1) and null hypothesis (H0); include direction if your variables are ordinal v. ordinal, ordinal v. ratio/interval, or ratio/interval v. ratio/interval.

· Provide visual analysis with appropriate evidence to support your argument:

· Crosstabulation: Compare two cells in one row with the biggest change in frequency column percent. Make sure to tell which two cells, in which row; give their column percentages; then subtract and use the difference between the two for your predication of significance and strength.

· ANOVA: Look at the Box and Whiskers diagram and compare the slope changes (purple lines, but don’t refer to them as purple lines!) between the highest mean and the lowest mean (red lines in the boxes, but don’t refer to them as red lines!) to make your prediction for significance. Then quantify the difference between the highest mean and the lowest mean as evidence for your prediction.

· OLS Regression: Look at the scatterplot to determine linearity, direction, and significance/strength.

· Report whether the relationship is significant or not with appropriate statistical test evidence:

· Crosstabulation: Chi-squared test result, degrees of freedom, and probability

· ANOVA: F test result and probability

· OLS Regression: F test result and probability

· Reject or fail to reject the null hypothesis based on the significance test result. If the relationship wasn’t significant you are done with your analysis – proceed to your conclusion.

· Report the measure of association/strength with appropriate statistical test evidence if the relationship was significant.

· Crosstabulation

· Lambda or Cramer’s V if nominal v. nominal

· Cramer’s V if nominal v. ordinal

· Tau-b if ordinal v. ordinal square (e.g., 3X3, 4X4, etc)

· Tau-c if ordinal v. ordinal rectangle (e.g., 3X4, 3X5, etc)

· ANOVA – Eta2

· OLS Regression: Pearson Correlation Coefficient (R)

· Provide the PRE or equivalent statement

· For PRE measures (Lambda, Tau-b, Tau-c, Eta2, and Pearson Correlation Coefficient (R2): Knowing [fill in unit of analysis such as “a person’s” or “a state’s” or a “country’s”] [fill in the independent variable] reduces the error in predicting [fill in the dependent variable] by [convert measure of association into a percentage] [add statistical test evidence for the measure of association such as R2=.423]. Remember, for the Pearson Correlation Coefficient (R) you must square the result to make the PRE statement (R2).

· For non-PRE measures (Cramer’s V): Approximately [Convert measure of association into a percentage] of the differences in [fill in the dependent variable] can be attributed to [fill in the independent variable] [add statistical test evidence for the measure of association such as V=.423].

· Confirm the direction of the relationship using the slope (b) (for OLS Regression only).

· Discuss the slope’s effect on the dependent variable using one unit of the independent variable (unstandardized slope or b=?) (for OLS Regression analysis only).

· Report the regression equation, filling in the variables and values (for OLS Regression analysis only).

· Provide appropriate tables and graphs with references in text (make sure data is aligned when you cut and paste).

· Final Conclusion

· Summarize your findings and compare and/or contrast them to the results you summarized in your lit review

· Provide your argument as to why you think the variables have this relationship. If the results from your significance test were not significant, you will need to greatly expand this portion to really analyze why your results weren’t significant when the results in your literature were significant (e.g., Was it something with your research design? Your data source? How you operationalized? Has the relationship changed and your results represent the true relationship? Etc.)

· Discuss shortcomings with your research and make recommendations for future researchers.

· Overall Appearance, Publication Style, Cites, and Grammar

· Does your paper look professional? Check:

· All pages are accounted for

· Nothing has shifted in your tables or paragraphs

· You have a cover page

· All pages except the cover have a page number in the upper right corner

· Stapled in the upper left corner

· Double-spaced throughout, including Works Cited page

· Minimized quotes – limited really to Intro to grab attention and perhaps Conclusion (i.e., you have put summaries, analysis, and evidence in your words and cited appropriately)

· All tables and figures have a consecutive number and a title and they are all referenced at the appropriate places in the text

· No first person wording – remove any “I” or “we” statements

· No normative wording – check for words such as shall, must, best, worst, good, bad, etc.

· Did you use MLA correctly? This is so important for this paper. I will return your paper to you and it will be considered late with a penalty for each day until it is returned in correct MLA format.

· Have you cited often and correctly using MLA? More is better. When in doubt, cite!

· Have you proofed your paper and read aloud, multiple times to check for typos, changes in verb tense (recommend that you use past tense throughout), poor grammar, and incorrect spelling? Get your friends to read it for you. Get your family to read it for you. Get your significant other to read it for you. I guarantee they will be impressed and give you useful feedback!

· Did you post an electronic copy of your paper, in both Word and a PDF, to Blackboard?

· Did you post your SPSS file (file with the .sav extension) to Blackboard?