Research Tips with Dr. Juliana

  SPSS output tables on a laptop with guidance on which results to report in research. Picture this: You have just finished cleaning your dataset. You navigate to the Analyze menu in SPSS, select your statistical test, click OK , and wait. A fraction of a second later, the SPSS Viewer window pops up. You scroll down. And down. And down. Suddenly, you are staring at twenty pages of output filled with descriptive statistics, reliability tables, correlation matrices, model summaries, ANOVA tables, coefficient blocks, residual plots, and collinearity diagnostics. A familiar wave of panic sets in, followed by a single nagging question: “Which of these tables am I supposed to put in my thesis?” If this sounds familiar, take a deep breath. You are experiencing SPSS output overload , a rite of passage for almost every postgraduate researcher. Here is the central lesson for today’s guide: SPSS gives you output; your job as a researcher is to decide what evidence belongs in your resea...

You run your reliability analysis in SPSS, and there it is: Cronbach's alpha = .68. Now you are wondering: “Is my questionnaire unreliable? Do I need to throw away the scale? Will my supervisor or journal reviewer reject my study?” Not so fast. A Cronbach's alpha below .70 deserves attention, but .70 is not a magical pass-or-fail line . The meaning of an alpha coefficient depends on the number of items, how closely the items are related, the nature of the construct, the purpose of the research, and the evidence that the scale is measuring a coherent construct. In this article, we will look at what Cronbach's alpha actually tells you, what α = .70 means, what to do when you obtain something like α = .68 , and the mistakes you should avoid when reporting reliability in a thesis or journal article. Quick answer: An alpha of .70 is commonly used as a rule-of-thumb benchmark for acceptable internal consistency, but it should not be treated as a universal cutoff. A...