Food for the Hungry Logo Barrier Analysis: A Tool for Improving Behavior Change Communication in Child Survival and Community Development Programs
Background Information
What is Barrier Analysis
How to Conduct Barrier Analysis
Defining the goal, behavior & target group
Developing the behavior question
Developing questions about determinants
Organizing the analysis sessions
Collecting field data
Organizing and analyzing results
Using the results

Photo of girls looking out a window
  How to Conduct Barrier Analysis

Before we begin to take you through the seven steps of Barrier Analysis, we want to begin with a brief description of the two approaches to this process: *

  1. Using focus groups (hereafter referred to as Option #1)
  2. Using individual interviews (hereafter Option #2).

Advantages of Focus Groups

  • Take less time: Doing two focus groups of 15 people will generally take about half a day (4 person-hours). Doing 60 or more 15-minute individual interviews (assuming several minutes between interviews for travel) will take at least two full days (about 15 person-hours).

  • Additional information: Focus groups allow you to ask questions that are not on your questionnaire to get a deeper and richer understanding of the situation in a particular area. When you are tabulating multiple questionnaires, these details are often not captured or not recorded. Many of the things that we found in the analysis done in the Dominican Republic, for example (see Session 8 above), would probably not have been captured if we had been using individual interviews.

  • Requires less "Doers": It is sometimes difficult to find 30 “Doers” of a particular behavior. In this case, it would probably be more appropriate to use Barrier Analysis through focus groups of Non-Doers. In that way, you can get richer details on barriers. Since you would not have a comparison group, there would be fewer benefits of a quantitative study.

Advantages of Individual Interviews

  • Requires less training: Using individual interviews generally requires less training and skill on the part of the people asking the questions. It is easier to administer a questionnaire for an individual interview than to facilitate and keep a rich and lively discussion going in a focus group.

  • Can quantitatively compare: Using individual interviews allows you to quantitatively compare the two groups, that is, to compare what portion of “Doers” have a given barrier or opinion vs. “Non-Doers.” However, the sample size needed to find differences between two groups that are not very different can be quite high. For example, you would need about 70 Doers and 70 Non-Doers to detect a difference of 20 percentage points between the two groups, and over 300 Doers and 300 Non-Doers to detect a difference of only 10 percentage points between the two groups. If small numbers of interviews are done (e.g., 30 for each group), even these quantitative results must be viewed with some skepticism. Only large differences (>30 percentage points) are generally meaningful when you have a sample size of 60 (30 Doers and 30 Non-Doers).

    For example, let’s say that you ask mothers “What are the advantages of exclusively breastfeeding?” Let’s say that you found that 8 of the 30 exclusively-breastfeeding women say that it helped avoid diarrhea, and 16 of the 30 Non-Doers—those not exclusively breastfeeding—said the same thing. You might want to say that since 27% of Doers and 53% of Non-Doers believe this, then that’s an important factor to take into account when designing your educational messages. However, the confidence interval for the 27% you found is actually 11-43%, so somewhere between 11% and 43% of the mothers actually believe that. For the Non-Doers, the confidence interval is 35-71%. Since these two confidence intervals overlap, there is a reasonable chance that the two proportions are actually the same. Even if you wanted to be 90% certain that there was a difference (instead of 95%), you would still have an overlap and could not show a true difference. You can overcome this shortcoming by doing a lot more interviews (e.g., 100 in each group), but that would probably be quite time consuming. One way around this problem is to include Barrier Analysis questions in larger surveys that you already have planned.

  • Less bias: Using individual interviews often leads to less bias since people do not hear the answers of others. Focus group participants are supposed to be selected in such a way that they do not know each other very well, but that is often hard to achieve in smaller communities. Sometimes leaders “insist” on being part of the group, as well. This can lead to a bias where most people in the focus group will “follow-the-leader” and give the same response as the strongest opinion leader in the group. Some people may not feel as comfortable saying some things in a focus group, either.

In the practicum (field trial), we will practice conducting Barrier Analysis both ways and then compare our results to see if our findings are substantially different.

Quick Quiz
Some of the advantages of using focus groups to do Barrier Analysis are that they take less time (than individual interviews), they provide a deeper and richer understanding of the situation, and fewer "Doers" are required.

True False


* Graciously provided by the Academy for Educational Development’s Change Project as part of their Doer/Non-Doer Analysis manual.

Next (Defining the goal, behavior & target group)

Food for the Hungry Logo

© November 2004

Home Site Map Contact Us