Hypotheses

Can you give an example of a directional and a nondirectional hypothesis?

Here are several examples:

Directional	Non-Directional
The more money a primary earner makes, the higher will be his or her sense of occupational achievement.	Income is related to a primary earner's sense of occupational achievement.
Including a diversity awareness program in the first-grade curriculum will increase the level of social acceptance among first graders.	Including a diversity awareness program in the first-grade curriculum will change the level of social acceptance among first graders.
Using a wholistic approach to teaching reading will result in greater reading comprehension than using a phonics-based approach.	Using a wholistic approach to teaching reading will result in a different amount of reading comprehension than using a phonics-based approach.

Is the following statement always true? A hypothesis is the answer to the research question that follows the literature review.

Almost always. There may be times when a researcher believes his or her theory or observations, in spite of the fact that this theory or these observations are not consistent with the findings of past studies or the observations of others. In these cases the researcher may propose an answer that contradicts the beliefs of other researchers who have studied a similar question.

Also remember that the hypothesis is the proposed answer to the research question. The primary purpose of the hypothesis is to provide a testable statement. If the proposed answer does not stand up to a rigorous experimental test, then we discard this hypothesis in favor of another.

When a hypothesis involves multiple explanatory variables, do hypotheses have to be written for each variable?

Except for the case of multivariate analysis (beyond the scope of this class), you do need to write separate hypotheses for multiple explanatory variables. For example, suppose your two explanatory variables are type of teaching (traditional or activity-based) and disability type (learning disabled or not learning disabled). In this case there would by hypotheses about the effect (or lack thereof, in the case of the null hypothesis) of type of teaching, hypotheses about the effect of disability, and another hypothesis about the interaction of these two. An interaction effect would be present if the effect of teaching depended on the disability level.

Does the way you write a statistical hypothesis depend on the statistics that are being used in the study?

Yes. Statistical hypotheses are hypotheses that can be directly tested using statistics. The specific statistical hypothesis will therefore depend on the statistics being used. For example, if the researcher is interested in comparing the mean achievement of two classes, then the statistical null hypothesis is that the population mean for one condition is the same as the population mean for the other condition. If a researcher is interested in correlating two measures, then the statistical null hypothesis is that the population correlation is zero.

URL http://edpsych.ed.sc.edu/seaman/edrm700/questions/hypotheses.htm

This page was last modified May 27, 2003
For comments or questions contact:
mseaman@sc.edu

The views and opinions expressed in this page are strictly those of the page author. The contents of this page have not been reviewed or approved by the University of South Carolina.