Question

Design a hypothetical study and discuss the design controls for the rival casual factors.

Answer 1

In experimental design, Rival Causal Factors refer to variables other than those being studied that are responsible for the outcome. In other words, rival causal factors are alternative interpretations, for the interpretations made by the observer.   

There are two categories of Rival Causal Factors - Internal (Variable within the study that interfere the study) or external (variables outside the study that may interfere)

To understand how we can control rival causal factors lets consider an experiment to find if stress management training is effective in reducing stress in a group sales officers.

Design controls to eliminate Rival Causal factors include

1) Use of control groups and random sampling: It is essential that test group (the group that undergoes treatment - in this case stress management training) and control group (that doesn't under go treatment - i.e. no training) are picked by random sampling. This ensures that there is equivalence in both the groups. It is essential that both the groups are similar in terms of age, experience etc. This is important as experience in sales, educational background, family backgrounds, salary scales could play an important role in the stress level of sales-force. It is essential that both groups are equal to rule of these other variables influencing the experiment results.

2) Use of Pre - and Post measurements: Pre-assessment of the variable (say blood pressure) in the test and and control group conducted should be conducted to find if the stress level in both the groups is identical or similar. The post assessment of groups after the test group undergoes the treatment thus, shows if there is any difference in stress level due to training. Pre measurements also further strengthen the premise that the two groups are equal at the start of the experiment and it is only the treatment (in this case the training) that is responsible for difference in post assessment results.

3) Instrumentation: It is essential that same instruments are used to measure the both the groups, for pre and post assessment in order to avoid variation due to use of different instruments. Thus in above example use of same method and equipment to measure the blood pressure in both the groups, as different equipment could be calibrated in slightly differently leading to variation.

4) Testing: This essentially arises due to foreknowledge of test in population that has been pre-tested. Prior exposure with test can lead to better results. In certain cases pre-tests are omitted to avoid this factor.

5) Placebo Effect: This is the tendency of subjects to react to a known stimulus in predicted manner.