Variables, measurement levels and hypotheses in your thesis

Many thesis topics involve variables. You have to identify which variables play a role, at what level you measure them, how you make them measurable, and what your hypotheses for the variables are. How do you do all of that properly? We'll briefly go over everything you need to consider in the process. Read on below and click through to the next articles for more information.

Variables are aspects of your research topic that you can measure and often manipulate. Examples include age, degree of self-confidence, the degree a person is completing, font size, or the degree of prior knowledge. 

There are all kinds of variables. In particular, there is an important difference between independent and dependent variables. The independent variable is the variable that you expect to affect the other variable. The dependent variable is the variable where this effect is noticeable. 

Suppose you are doing research on if the number of  hours spent studying influences students' exam results. Then, the number of hours spent studying is the independent variable. Exam results are the dependent variable. 

In our article on variables, we provide more examples and explanations of the various types. 

You can measure variables at multiple levels. For example, the variable "education level" consists of categories that you cannot express in numbers. The variable "body weight," on the other hand, you can express in numbers which also means you can calculate an average for it. All of this refers to the level of measurement. 

There are four levels of measurement in statistics:

• Nominal variables: variables that fall into categories and have no ranking (such as gender or the transportation people use to get to work). 

• Ordinal variables: variables that fall into categories and are ranked, but that don’t have equal intervals (as with unequal age categories or education level).

• Interval variables: variables that you can order and are spaced by equal intervals (such as for equal age ranges: 25-35 years, 35-45 years, etc.).

• Ratio variables: variables that you can order, that have equal intervals, and have a zero point (such as age or number of years of work experience). 

How you measure a variable also affects the statistical analyses you can and cannot do.

Read our article on measurement levels for more information, examples and explanations.

You can only measure many variables after you have made them measurable. Think of a variable like "degree of self-confidence." There are many ways to measure this: having respondents fill out a questionnaire, self-assessing respondents based on observations, or using an existing measurement method for self-confidence. Either way, you can only measure this variable if you choose an appropriate measurement method for it. This is called operationalization. 

How exactly does operationalization work and how you do it? You can read that in our article on operationalizing. 

Once you have your variables and the research question clear, you formulate hypotheses. This means that, based on literature review, you predictthe answer to the research question(s). You will then test those hypotheses in your research. 

Read in our article on hypotheses  to learn how to build a good hypothesis in 3 steps, including several sample hypotheses.

Setting up research for your thesis is an important but often tricky thing to do. Where do you start? What research method do you need to use? We would like to help you with this through a number of our useful articles. For example, read our tips for a good research plan, our articles on research methods, and our article on reliability and validity. 

Do you have doubts about the text of your thesis? Are you afraid, for example, that your thesis is not well put together in terms of structure, or that the thread throughout it is not entirely clear? Check out our thesis review service. Our thesis editors are happy to give you personal feedback, even within 24 hours notice!