EMMLOGE

QUESTIONNAIRE CONSTRUCTION

The construction of a questionnaire can be divided into 6 main steps:

Topic selection and writing the research problem: The research problem would indicate the nature of the problem for which answers are needed. To write a good research problem, the scholar or researcher should learn as much as possible about the subject matter before setting out to write the questions.
The researcher should determine the type of questionnaire needed. For e.g. the type of questions will depend on the mode of administering the questionnaire as well as the target population and the type of analysis to be done.
A set of variables on the topic are set out. The variables are used to do a draft set of questions.
A re-examination or revision of the questions that have been drafted is done. The re-examination helps to remove bias and poor question format. It also helps to remove technical faults. The questionnaire is given an introduction and to assure respondents of anonymity and confidentiality, no name of respondents is asked. The questionnaire should rather have the name of interviewer. i.e. the person asking the questions and the questionnaire should have QID (questionnaire identification number) When in doubt ask for professional help from SPSS programmers.

GUIDE TO DESIGN QUESTIONNAIRE DESIGN

SEE QUESTIONNAIRE

ADVANTAGES OF THE MAIL QUESTIONNAIRE

The mail questionnaire involves the printing of the questionnaire and mailing them to selected addresses. It has certain advantages compared to the face to face interview schedule

1. Mail questionnaire is cheaper, impersonal and can be administered for a large sample

2. It eliminates the time consuming task of training field staff to administer the questionnaire.

It only requires self addressed envelopes with stamps for the return of the questionnaire.

3. The respondents may answer questions more frankly since anonymity is assured.

4. It reduces errors that may result from interviewer bias.

5. The questionnaire can be answered at the convenience of the respondent because it gives the respondent time to think through the answer to give.

DISADVANTAGES

1. The response rate is usually low between 10-50%.

2. It requires a higher level of education from the respondents. It is therefore not recommended for collecting data from rural population that tend to be largely illiterate.

3. The answers have to be accepted as final. There is no opportunity to probe beyond the given answer to clarify ambiguity or even to note the behavior of the respondents in answering the questions

4. The researcher cannot be sure that the right person would complete the questionnaire.

5. The respondent can see all the questions before answering any of them. It is therefore difficult to check on the honesty and reliability of the data.

*correlation

x- Independent variable

y- Dependent variable

X²TEST (CHI SQUARE TEST)

BASIC DECISION MAKING TEST

Chi square analysis involves the application of correlation to qualitative or non-quantitative data. It deals with analyzing the relationship between 2 variables which are qualitative in nature.

The computations assume that there is no relationship between the two variables.

On the basis of this assumption, critical or expected frequencies are computed. The expected frequencies are compared with the original data which is also called the observed frequencies and the greater the difference between the expected and observed frequencies the stronger the relationship between the 2 variables.

The measure of the discrepancy between the observed and expected frequencies is called Chi square.

It is defined as follows:

O_F=observed frequency

E_F=Expected frequency

NON-PARAMETERIC STATS=CHI SQUARE

SOME USES OF THE CHI SQUARE

The Chi Square test offers the following uses for researchers

It is used to determine whether a particular frequency distribution could develop by chance from a normally distributed population
Chi Square is used to investigate the relationship between attributes or traits of individuals which can be classified into 2 or more categories i.e. it is used to test qualitative variables.
Chi Square is used to do a test of significance of the agreement between the observed and expected results or variable

Researchers conducted a survey about the relationship between marital status and IQ

The following table is the data compiled from the survey

Determine whether there is a relationship or association between marital status and IQ

Do the test at 5 % level

· If not stated assume at 5 % level.

Whenever you do a test of significance you will come to a certain conclusion called confidence level i.e. you are leaving a margin of error.

Solution

Step I

Set Null and alternative hypotheses

H_O=Null hypotheses

(Hypotheses of no relationship between 2 variables)

H_A=Alternative Hypotheses