Skip to Main Content

CAPSTONE PROJECT: Analysis of Qualitative Data


I      Intro
Majority of what it is covered comes from an eBook in our collection. Chapters 5 and 10 are good reads. It is a book for Market Research, so not exactly a “church” or “ministry” book, but put you and your topic in as “clients” and the book makes more sense.
Analysis is the deep interpretation of the data gathered. It is not simplistic. It is not skimming or rushed. It is not superficial or based on impressions. It is thorough, detailed, and takes time.
Analysis is not just based on summaries of the data, but going back to the original data, looking for anomalies or patterns that don’t come out in a simple summary. In some ways, the real data is not the hi/low and median, but the odd ball numbers that skew the data.
II      Big Ideas in Analysis
            A Patterns / Clusters / Themes
Patterns and themes can often be seen while looking at a spreadsheet. Sort the data by columns and see the strong correlations between answers. Often you will note that many people that answered one way on one question will answer the same on another question. Those strong correlations are the themes that you may be looking for. Some may surprise you.
            B Discontinuities / odd data
                  1 when someone says yes to one thing and no to another and that is just wrong
Just as you notice patterns you can sometimes spot discontinuities; places were the answer is unexpected. I mentioned sorting a spreadsheet and looking for correlations. Here, you are looking for the outlier, the odd data. It might look like a mistake; how could they answer that way, when they answered such and such on this question. In other words, look back at those patterns and then look for that one or two responses that seem to go the opposite direction. As bad examples, the tree hugging Republican or the gun toting Democrat.
                  2 the outlier data, the 1 on the scale when everyone else is an  8
Another aspect of this is the outlier data. We often want to ignore the outliers, the odd numbers that mess up the curve or lower the median. Those outliers are (most of the time) real opinions and they need to be accounted for. Just because some data is opposed to the majority, doesn’t mean it is invalid. True, you may have a poorly worded question and maybe it was misunderstood, but this may be someone’s real opinion. Look at the data and try to imagine why someone answered this way. Will you need to account for this in your findings? Maybe, if there is enough or if the opinion is strong enough, or the outliers create their own pattern.
            ; C Generalization (positive/negative, or overall impression)
You also need to analyze general ideas, not just patterns. Here you are looking at the strength of convictions and is there an overall impression. When people choose 10, not 9s or 8s on something, they are not only stating agreement, but strength of conviction. A 1 is basically a screaming 3; both are disagreements, but the one is stronger.
In many ways, this is your quick read of the data, but it is after looking at the patterns and such. If you had one sentence to describe the data in relationship to your problem, this is how you would say it. Again, here is not the conclusion, but your generalization; “the data seems to say that most people feel / think that we ought to…”
III      Relationships to Look at
            A patterns/clusters/themes and personal data – groups and their identifiable ideas
You have already looked at the patterns and such, but now go back and look at the “who” of the pattern. Is there a dominate type of person that created the pattern? Women in their 40s-60s feel one way while the males in their 20-30s feel the exact opposite. Look for the personal data in the patterns. You may find you have different blocs of people that create their own patterns. It is expected that if you have a large enough data pool that you should be able to spot patterns and that those patterns may follow certain demographics. You must pay attention to that data, it may help solve your problem. It may help you figure out how to promote the idea to different groups. It may help you figure out how to motivate the different groups to work on a common outcome. Some out of duty and faith and others out of pride or sense of ownership.
            B discontinuities and personal data
You again look back at the patterns and now you look at the discontinuities and compare to personal data. Are these odd ball responses from people of influence? Are these fluke answers? Possible misunderstandings? OR Could it be that this outlier is actually the most important opinion of all?
You will have looked at that patterns and the personal data and will have made connections; this is a continuation of that same idea. But now you are looking at the minority voice. Why did this person or people group answer this way?
            C generalizations and personal data
As you continue working with the personal data and its relevance you may become aware that you have more than one dominant group. This is often a source of conflict; multiple groups with slightly differing agendas. In my church right now, we have a lot of people that are pro-missions. The struggle is that we are talking across each other as one group is talking about foreign missions and another is talking about cross cultural ministry here in the country. They both love people of differing cultures, but one wants our focus over there and the other wants it here in the neighborhood.

Life Pacific University Alumni Library | 1100 W. Covina Blvd | San Dimas, CA 91773 | Ph: (909) 706-3009 | Email:

Return to Life Pacific University homepage | Log in to LPU email