Hypohtesis Testing

Steps to hypothesis testing.

Hypothesis testing can be summarize in a six step procedure:

• Establish a null hypothesis as well as the alternative hypothesis.
• It is a one-tailed test of significance if the alternative hypothesis states the direction of no difference in the other hand if no direction or difference is given, it is a two-tailed test.
• Choose the statistical test on the basis of the assumption about the population distribution and measurement level.
• The form of the data can be a factor, in that case you will typically chose the test that has the greatest power efficiency or ability to reduce decision errors.
• Select the desired level of confidence.
• =0.5 is the most frequently used level, many others are also used.
• Compute the actual test value of the data.
• After the data is collected, use the formula for the appropriate significance test to obtain the calculated value. The computation results come from a software program.
• Obtain the critical test value, usually by referring to a table for the appropriate type of distribution.
• The critical value is the criterion that defines the region of rejection from the region of acceptance of the null hypothesis.
• Interpret the result by comparing the actual test value with the critical test value.
• For most test if the calculated value is larger than the critical value, we reject the null hypothesis and conclude that the alternative hypothesis is supported. If the critical value is larger, we conclude we have failed to reject the null.

 Hypothesis testing is crucial because it helps us to reason from evidence found in the sample  to conclusions. The virtues of this steps also allow us to to determine the accuracy of the hypotheses due to the fact that you have collected a sample of data.

Business Research Methods, Eleventh Edition. D. Cooper/P. Schindler

Pretest Survey Instruments

Why is desirable to pretest survey instruments?

Researchers should pretest survey instruments because:

• It helps to discover ways to increase participant interest
• Increases the likelihood that participants will remain engaged to the completion of the survey
• It helps discover question content, wording and sequencing problems
• Discovers target question groups where researcher training is needed
• Helps explore different ways to improve the overall quality of survey data.

The information that a researcher can secure by pretesting is the participant's reaction to the questions. Also, the researcher can secure errors in the questionnaire and fixed it along the way.

The best wording for a questionnaire can be found by asking your self the following questions:

1. Is the question stated in terms of a shared vocabulary?
2. Does the question contain vocabulary with a single meaning?
3. Does the question contain unsupported or misleading assumptions?
4. Does the question contain biased wording?
5. Is the question correctly personalized?
6. Are adequate alternatives presented withing the question?

The researcher must know who the target participants are going to be and what kind of answers is looking for in order to create a good set of questions.The vocabulary use in the questionnaire must be understood by both the researcher and the participant. The vocabulary can be found by using simple English words and phrases. You should also remember that questions should not be only constructed on words, you can incorporated visual images as part of the questioning process.

Business Research Methods, Eleventh Edition. D.Copper/P. Schindler

Four major sources of measurement error.

Four major error sources:

• The Respondent
• The Situation
• The Measurer
• The Data Collection Instrument

The Respondent

When it comes to a face to face interview the researcher can encounter  participants that don't want to give a strong negative or positive opinion.  Also, the participants may have no knowledge of the subject with no intentions of revealing that to the researcher, which can cause the data collected to be full with guesses or assumptions.

The Situation

In a face to face interview the situation factor can affect the data in different ways. First, if someone is present during the interview this person can interrupt, distract or just by being there can change the participants answers. Second, if the participant is afraid that the interview responses are not going to be kept anonymous the may be reluctant to give the right answer. Finally, participants in face to face interviews give less elaborate answers compare to the home interviews.

The Measurer

 The interviewer can change responses by paraphrasing, rewording or reordering the questions, which can cause a major change in the responses by the participants. Also, not concentrating by checking the wrong response or not recording the complete responses can distort the findings.

The Instrument

If the researcher instruments to conduct the interview can cause distortion in two major ways. First, it can cause confusion to the participant by using complex words beyond the participant comprehension. The use of mechanical defect tools, not waiting or giving enough space for response, and poor printing can cause the final data findings. Second, poor selection from the universe of content items. The interviewer can create a bad survey by not including more broad questions about the entire subject or company that they are trying to investigate. Even when the general issue or subject are studied, the questions may not cover enough aspects of each area of concern.