Every week, I receive between two and five emails asking about research questions, most of which I do not answer because I don’t have time, and because I have already answered them (or similar questions) repeatedly in this blog. When I do reply, it is usually because an email provides the affordance for a teaching point, and the message that follows provides no fewer than three:
- What is the difference between Dependent and Independent variables?
- Must we really use SPSS to do statistics?
- What are some good norms for requesting assistance?
So here goes:
Hello! I am conducting a survey on the “Incidence of Credit Management Policy on the Profitability of a Commercial Bank” . My questionnaire has 12 items where 33 respondents are to select their agreement to the topic: (1) strongly disagree – (5) Strongly agree. I have tried to find out which variable is dependent or independent but in vain. I have succeeded in interpreting each item separately, but with mean values and Standard Deviation which you don’t recommend for ordinal data. In one word, I don’t see steps to calculating Pearson correlation for that is my objective as an ending product of my survey. I need help and it is very urgent.
Dependent and Independent variables
The short answer is that “profitability” is the dependent variable, and “policy” is the independent one.
Now for a longer and more useful one. A hypothesis usually connects two or more variables, which I will call variables of interest. Of these, independent variables are variables that do not change as the result of some other variable of interest. For instance, if we are researching family income and educational outcomes, family income is an independent variable, because families do not earn more money if their children do better at school. Dependent variables are variables that change if there is a change in another variable of interest. For instance, I might hypothesise that children from low-earning families struggle at school, compared to children from more affluent backgrounds. School performance, in this hypothesis, is a dependent variable.
More to read:
- Cohen, L., Manion, L., & Morrison, K. (2000). Research methods in education (5th ed.). New York: Routledge. Chapter 10
- Muijs, D. (2004). Doing quantitative research in education with SPSS. London: SAGE Publications. pp. 121-122
SPSS is a very powerful statistical package, and it is very helpful for students to be familiar with it. However, many statistical procedures (including estimating averages, means and medians, standard deviations and IQRs, and correlations) can be easily done using widely available software like MS-Excel, OpenOffice and GoogleDocs. All these software packages contain a function called PEARSON or SPEARMAN, and help pages explaining how these can be used. Typically, you type =PEARSON(array1;array2) in a cell, substituting ‘array1’ and ‘array2’ with selections of the cells where the dependent and independent variables are listed. It makes no difference which variable is selected first.
More to explore:
- Calculating Pearson’s r with MS-Excel, Open Office, and Google Docs.
- Four ways to calculate Spearman’s rank correlation coefficient.
Meeting Miss Manners
I am aware that politeness norms are culture-specific, so I will not interpret “I need help and it is very urgent” as being indicative of an entirely undeserved sense of entitlement, nor shall I infer that the person who sent the email was inconsiderate to the fact their (urgent) needs should take precedence over whatever else might be salient in my own schedule. However, in the interest of developing their intercultural skills, I must point out that occasional use of the word ‘please‘ can make a person appear much more civil in English-speaking societies. Additional insights into polite communication can be found here.
Featured image: Michael Kwan [CC BY-NC-ND]