MARCH/APRIL All-Star Tip of the Month: Conducting Quantitative Surveys

Here is a combined March & April Tip of the Month (I know you were wondering where the March tip was!) on conducting Quantitative surveys. A big THANK YOU to Bogii for providing us with 8 important tips based on her recent experiences.   Bogii just conducted her full survey in three new sites along with a short survey on a new behavior at 4 existing sites from last year’s campaign. For the new sites, Bogii removed section 3 on “trusted sources and media preferences” because she will not be changing her choice of materials or creative designs for those sites (she will use her existing materials). She also removed questions that were not relevant to her preliminary objectives – this resulted in a survey that only took 25 minutes per person to implement. For the existing sites, Bogii created a very short survey (13 questions total) that included some basic demographic information for filtering and focused only on knowledge, attitude, IC, and behavior. Bogii is still entering her results into SurveyPro, but has sent us the following tips based on her recent survey experience:   Be careful who you pick as your enumerators. Do not use official inspectors (or government officials, policemen, etc.) as enumerators. Many people know who they are and will not answer honestly. Training your enumerators is a very important step, even if it takes some extra time. Check your enumerators’ perceptions of the questions and what you are asking. Explain best approaches for collecting responses from residents. Avoid national holidays during your survey collection time. People likely won’t take the time to answer your questions on those days and planning around the holiday will prevent your timeline from getting delayed. When visiting sites for surveys, plan ahead to see what else you can accomplish while traveling. Bogii met with Government organizations during her survey visits to discuss future campaign plans and secure their support for her activities. She even met with one school eco-club who has agreed to help her plan and design a May event at their site. If you want respondents to be in the right ‘frame of mind’ when answering your questions, then explain a little bit about the environmental and economic issues your questionnaire will include so they understand why you are asking questions. Be considerate of the time of day you are collecting surveys – don’t conduct them during peak times of family’s work schedules. Bogii found that the best times for her enumerators were 11-13pm (lunchtime) and 15-17pm (afternoon). When in a respondent’s home, take notice of the surroundings and the atmosphere in the house (sad, angry, happy, etc.) This can provide you with some informal qualitative research and help to understand some of their survey responses too. Be careful of herders’ dog in countryside, very scary.   I also want to include a question and answer about how to select respondents who share a home that was recently asked by one of our partners in the field. I hope this will be helpful as well:   QUESTION: We know that it’s important to talk to both male and female heads of household, but what is a practical way to do this when entering a respondent’s home and both husband and wife are present? Do you toss a coin, alternate, ask each interviewer to talk to 15 males and 15 females? Should we be considering other factors such as age, or head of household/non head of household?   ANSWER: If you are going for a completely random sample, then determining # of males and females you need at each site to get a representative sample will help guide this (if total population has 60% males/40% females, then you would want to replicate that in each town/village with your sample sizes). Then you can ask your interviewers to pick one member of the household at random in order to achieve the #’s they need.  If you have a very specific target audience – say, male hunters, then you could skew the numbers so you interview more male hunters (and more males period) to get more responses from that group. Then you could advise your interviewers to pick the males if they’re home. Alternatively, if your target audience isn’t defined by demographics – for example, say you need to interview the person who makes the food purchasing decisions in the house – then that’s a lead-in question the interviewer could ask when entering the house and then only interview that person. The things to watch out for are cultures where even if the woman is making the decisions, the man will answer that HE is the decision maker. If that’s the case in your site – then doing a more random sample (coin toss, alternate households, etc.) is probably a good way around that, or default to point #1. Because of the above, you want to be sure you don’t interview husband or wife in front of the other person. That will undoubtedly produce bias in your survey. Rare generally doesn’t survey children under the age of 14 unless they’re a key target audience.