1. Gain enough experience with SPSS and its datasets and procedures that you can complete the analyses for your dissertation with little or no assistance.
2. Practice for the statistics Ph.D. qualifying exam. The process below is a description of how I prepare each qualifying exam. If you learn the process it will help you pass the exam.
3. Realize that the results of statistical analyses depend on the continuous or categorical coding of the variables, and the type of analysis that is done. Results that are "real" are found regardless of the type of analysis. Results found in one type of analysis but not in another are probably artifacts of the analysis.
Preliminary Steps:
1. Obtain a dataset that has most of the variables that are relevant for your study (see the literature). The dependent variable in this dataset should be continuous.
2. Begin creating a SPSS syntax file that is used to analyze the data.
3. Examine all of the relevant variables, insure that they vary, and specify missing value codes.
4. If a variable is measured by three or more indicators, create an index from the indicators.
5. Create categorical and continuous versions of your dependent variable and your most important independent variable(s).
Introduction:
The paper begins with one or more introductory paragraphs that indicate the topic, the research problem, and the purpose of the study. These paragraphs are interesting and insure that the reader wants to read the results.
Research Question:
In a paragraph of two or more sentences, present the research question. This is a strongly worded research question about the effect of one (independent) concept on a second (dependent) concept, controlling for one or more other (control or confounding) concepts.
If you are not sure that your independent variable is causally related to your dependent variable, then you may write a weaker research question asking about the relationship between one (independent) concept on a second (dependent) concept, controlling for other concepts.
Literature Review (3 pages):
Your literature review should in a concise and objective way summarize the relevant, current (within the last five years) literature that presents results of research about your topic. This review should use citations according to the style of the American Psychological Association (version 6).
Definition of the word "relevant": I encourage you to, as the first step, find and list all current and classic journal articles that deal with your subject. For example, a student writing about the effect of nutrition on athletic performance should find all works on nutrition and athletic performance that have been published within the last five years or that are considered to be classic works. Then, from this long bibliography, you are to select your references for your proposal carefully so that (1) the most directly applicable works are referenced and discussed in the review of the literature, and (2) no indirectly applicable but important work on the topic is omitted. This second criterion is stressed so that you will not find that your manuscript is rejected because a reviewer believes that some unreferenced article is important. Thus, "relevant" literature is that set of works that reviewers of journal articles, who are experts on your topic, are familiar with.
Your references should be articles from refereed scholarly journals, theses, and dissertations. You may also cite published reports from by public, non-profit, or governmental organizations (such as the Centers for Disease Control and Prevention, or the National Collegiate Athletic Association).
Read the articles and look at their hypotheses, statistical tables, and results sections. Keep track of which variables they used to specify their models, and which variables were kept in the models because of significance and which variables (significant or not) were kept in the models because they were theoretically specified to be important.
The literature review should be organized to identify all variables and interactions between independent variables that are necessary to specify your analytic model. Begin the review with the literature about your dependent variable. Then provide the literature on the one or two most important independent variables, specifying the extent to which they have effects on (or are related to) your dependent variable. After these two major sections, relatively short sections will review the literature for your less important independent variables (the controls) and their effects on the dependent variable.
The final paragraph of the literature review is a summary of two to three sentences that restate the research question with all relevant control variables.