Research Hypotheses and Hypothesis Testing

To begin an experiment in human-robot interaction, researchers usually formulate a research hypothesis that can be tested empirically. Unlike a theory, a hypothesis is a more precise and focused statement that can be tested by a single experiment (Rosenthal and Rosnow, 2008). The research hypothesis is a critical component of the experiment, serving as the basis for statistical significance testing. 

Typically, an experiment includes at least one null hypothesis and one alternative hypothesis. The null hypothesis usually assumes that there is no difference between the experimental treatments, while the alternative hypothesis presents a mutually exclusive statement to the null hypothesis. The objective of the experiment is to find statistical evidence to reject or nullify the null hypothesis and support the alternative hypothesis (Lazar, 2017). Some experiments may involve multiple pairs of null and alternative hypotheses. 

It is possible to test multiple hypotheses in a single experiment, but it is generally recommended to avoid testing too many hypotheses in one experiment. As the number of hypotheses to be tested increases, so do the factors that need to be controlled and the variables that need to be measured, making the experiment more complex and increasing the risk of design flaws. 

Hypothesis testing is a type of statistical analysis in which you put your assumptions about a population parameter to the test. It is used to estimate the relationship between 2 statistical variables. Usually, the hypothesis testing goes through the following steps: 

• Define the null and alternate hypothesis. 
• Collect data. 
• Perform the appropriate statistical test(s). 
• Decide whether to reject or fail to reject your null hypothesis. 
• Present the results. 

To start the analysis on a human-robot interaction experiment, the researcher establishes a null hypothesis based on the research question or problem that they are trying to answer. Depending on the question, the null may be identified differently. For example, if the question is simply whether an effect exists (e.g., does X influence Y?) the null hypothesis could be H0: X = 0. If the question is instead, is X the same as Y, the H0 would be X = Y. If it is that the effect of X on Y is positive, H0 would be X > 0. If the resulting analysis shows an effect that is statistically significantly different from zero, the null can be rejected. An alternative hypothesis is a direct contradiction of a null hypothesis. This means that if one of the two hypotheses is true, the other is false. 

For further studying watch this video: 

References 

Lazar, J. , Feng, J. H., Hochheiser, H.  (2017), Research methods in human-computer interaction: Morgan Kaufmann, 2017. 

Rosenthal, R., Rosnow, R., 2008. Essentials of Behavioral Research: Methods and Data Analysis, third ed. McGraw Hill, Boston, MA.