Analysis & Evaluation

If you did an Online Survey the program will do this for you, but you can still keep reading to find out what to do with these tables. Your online survey program will probably also provide you with the possibility to correlate results and calculate percentages, but it’s always good to understand what the program is doing or able to do.

If you are familiar with Excel transfer the results to an Excel chart. If not, use any kind of other program (e.g. Word) which supports tables.

Each question gets a table like this:

Education Question 5

The tables for question 1-9 only give an overview of your target group: You can check, if your sample is similar to the population as a whole. Here, 20% of the women in your sample are illiterate. If the female illiteracy rate in the population is significantly higher or lower you have to be careful when drawing conclusions from questions 10-15. You will have to do correlations first before you can say something about the overall FGM-rate.

If you find that your sample is similar to the population (or to the target group you chose, e.g. students) then you already have reliable results from questions 10-15.

FGM-Under Question 10

Your result is: 75% of the target group (e.g. women in city Z) have undergone FGM – IF and only if your sample matches the population. If your sample does not match the population you need to phrase this result more carefully. You can say: In an unrepresentative study among women in city Z 75% of FGM prevalence was found.

Correlation

You can check now how often an indicator (age, education etc.) correlates with the women having undergone FGM, their plan to circumcise their daughter and their opinion about the continuation of the practice. Vice Versa you will also see which indicators are most common in women who did not undergo FGM.

This will tell you more about the spread of FGM in different segments of society. In this example you can see if the educational level influences the probability that a woman has undergone FGM.

FGM-Under/Education

Avoid mistakes: When you read this table you always have to keep in mind that not all groups (illiterate, secondary etc.) are equally big. Do not just look on the left column and assume – as in this case – that FGM is most common among graduates of secondary school.

You have to calculate percentages now. Your results must always be in percentage but unlike in the simple one-question-tables the number in the table does not equal the percentage here.

If we want to know how high the prevalence of FGM among secondary school graduates is, we need to know what percentage is 20 of a total of 30.

Here is the simple formular:

20/30 x 100 = 66.66%

So, in this example 66.7% (you can always round) of secondary school graduates have undergone FGM. If you do calculations for the other educational levels you will find that in this example FGM is not most common among secondary school graduates.

Put your results in an extra table!

FGM-Under/Education percentage

Now, you have clear results on a glance and you can draw a conclusion: FGM is more common among women with lower education.

Avoid wrong conclusions: If a group is too small to be significant, you have no result. If you interviewed less than 10 women who are illiterate, the result could be mere coincidence. This is especially true, if their answers don’t give a clear picture. In our example there are less than 10 women who have had primary education. To draw a conclusion from their answers is problematic. However, the picture with 8 saying Yes and 1 saying No is very clear. If it was 6 saying Yes and 3 saying No, you can NOT use this result. OR you can combine groups: Take primary education and illiterate into one group, then you have 31 – any group bigger than 30 can definitely give you reliable results.

You should correlate all questions 1-9 with questions 10-15 and see if they give you significant results. Some may not and you may not use them later on when you publish your survey, but you should have checked.

You can also correlate questions 10-15 with each other. This may yield interesting results as well.

For example take question 12 “Why did they practice FGM on you?” and questions 15 “Do you think FGM should be continued?”

FGM Why/ FGM Continue

If you calculate the percentage in this example you will see that 50% of those who were cut because of tradition want to continue the practice. Of those who were cut because of religion 83% want to continue. So, you can conclude that religion is a much stronger reason for the continuation.

If you have done all the calculations you find useful you can read now how to write up your survey results as a paper which can be published in the section: publication.