Wednesday, March 12, 2014

c.4: Data Analyses: Quantitative


Today we will take a look at the results section for quantitative studies. Start out by reporting descriptive statistics "that appropriately characterize the sample." What does this mean? Look at frequencies for your demographics, such as gender, marital status, etc. For continuous variables (not in categories) you will need to compute the means and standard deviations or standard errors (check with your committee as to which they prefer). An example of such a variable is age, the convention is to give these stats like this (M = 43 yr., SD = 5.2). You will also discuss any total scores or subscores that you may have calculated and their distribution. 

The next step is to discuss and evaluate statistical assumptions as appropriate to the study. All statistical tests have specific assumptions that must be considered (see Pallant, 2013, for an in-depth discussion of them).  Let's take as an example, the assumptions for parametric tests (e.g., t-tests, analysis of variance): using an interval or ratio scale of measurement, random sampling, independence of observations (no measurement is influenced by another), a normal distribution, and homogeneity of variance (samples have similar variances). There are techniques to check these assumptions, and you would discuss in this section which ones you used and the results. 

Next, you report your findings, organized by research questions and/or hypotheses. Include the exact statistics and associated probability values (some examples: t(32)=3.1, p < .01; r(N=45)= .16, p > .05). A reminder- if the probability is < .05 (less than), it is considered significant; if it is > than.05 (greater than) it is not significant. You should include confidence intervals around the statistics, as appropriate (check with your committee). Include effect sizes, as appropriate (e.g., R2,; check with your committee as to what they prefer).  

If you had multiple conditions, you may need to do post-hoc tests. Report the type and results of post-hoc analyses.  You may have additional statistical tests of hypotheses that emerged from the analysis of main hypotheses and you will need to report those.

Finally, you may wish to clarify your results with tables and figures, include those as specified in the APA manual. There is very specific formatting for these- so check it out in the manual.  

Next time we will talk about c.4: Data Analyses: Qualitative. Do you have an issue or a question that you would like me to discuss in a future post? Would you like to be a guest writer? Send me your ideas! leann.stadtlander@waldenu.edu  

Pallant, J. (2013). The SPSS Survival Manual, 5th edition. Open University Press. 

1 comment:

  1. Dr. Stadtlander
    You mentioned “effect sizes, as appropriate (e.g., R2,; check with your committee as to what they prefer).” I have three questions.
    My first question is if my research is conducting a multiple regression analysis I do not need to calculate the sample size. Is this correct?
    The second question is when I perform sample size calculation, does the mean value indicate the value of dependent variable? Does the mean value of a population is based on a pilot study or other researchers' results? how to know the mean value for my population of CPAP adherence?
    Final question is when we report R2, should we also report adjusted R2 or both ?
    Thanks
    Molly

    ReplyDelete