Kiva Friends - Better polling needed

Sorry I'm coming in late on this discussion but better late than never (unless I'm "too late" on this one - sorry again!)

I'm not a "survey expert" but I have completed a dozen or so statistics and methodology classes (yikes!), including survey research, so I have maybe one & a half cents to contribute to the discussion. I do think the idea of conducting surveys is a good one - but they need to be carefully designed and kept SHORT to be effective - if a piece of data isn't relevant, don't ask for it. That is, if there isn't going to be a multivariate analysis correlating attitudes, number of loans or other survey items with age, gender, location or any other criteria, or if it isn't going to produce USEFUL information, don't ask.

Surveys should start with a well-defined research question - something you are trying to determine and make sure it is free of bias before going further. Then analyze each question to see how it relates to the research issue.

Instead of "Country of Citizenship" or "Ethnicity," why not "Country of origin" if that's relevant - people who have moved to another country have a different perspective than those who have lived only one place. It might be just as interesting to know about, let's say, Cambodians living in California as a Californian living in Cambodia. However, the main issue is whether this is going to be USEFUL rather than "interesting" and how will this data be used in the overall research design?

I would suggest that if you want relevant and useful data that is VALID, be sure to eliminate all bias in questions. For instance, "I am concerned about lending to people who do not seem to be genuinely disadvantaged" is a fairly leading question - I made me think "Well, duh, none of us want to loan to someone who wants to refinish the teak in their yacht" but the term "genuinely disadvantaged" is very relative and cannot be determined simply by looking at the country they are located, in most cases.

For instance, a lot of people seem to think capital is readily available in the United States - IT IS NOT, especially for small-time borrowers!! It can actually be easier to get a loan for millions of dollars than it is for $3500, although most individuals are finding that it's just not easy to get loans at all these days, regardless of what may have been the case in the past. So is the research issue proximity to wealthy people or personal circumstances relative to the surroundings? I have no way of knowing whether someone in Peru is the "richest person in town" or one of the poorest when I see their loan on Kiva - so is the real question one of determining individual poverty level or is it simply an anti-US bias? All of that needs to be worked out as part of the research design before starting on the survey!!

There also seems to be the perception that housing, medical care, food and economic opportunities abound here, which is the truth for some but absolutely not for all US residents. Right now we are in the middle of a terrible recession and many who never experienced hardship before have become desperate and we can't judge others by a superficial perception. So are we trying to determine whether an unemployed person who is living in their recent year Honda Odyssey van with 4 kids after they lost their home to foreclosure because they were unable to find a job after they were laid off might be less needy than someone in Uganda who has a home, food and clothing but wants to buy more cows? It's not something that ANY of us can determine by superficial examination so questions like that seem designed to promote an agenda rather than actually discover facts. A good research design needs to eliminate preconceived ideas or assumptions to produce valid results.

The same can be said about many of the "concerns" questions that have been mentioned - each of us has our own idea of what might be a concern! Whether it's type of business, location, amount of loan or something else about how it's structured, there are probably a million reasons why each of us might choose to skip over another loan. If we REALLY want to discover a truth, we need to be very careful about how the survey questions are structured and whether they will reveal the answer to the primary research question. It may take MANY surveys to gather information and still not be enough to result in legitimate answers. For instance, do we want to include questions about whether Kiva loans should be used to fund pubs? Or polygamists? Do we want to ask about any of the million other things that are "of concern" to "someone" out there!! [No intent to offend anyone by picking those two examples - I just wanted to point out that we need to know how much truth we are searching for when we design the research project - so do we want to know ALL concerns lenders have or are we simply trying to get more information about a select few?]

A good survey should also have a very clear idea of how the data is going to be analyzed - it's not an easy task to accomplish that, which is why open-ended questions like "how do you define poverty" could result in a nightmare task of analysis. Maybe a discussion board "focus group" could generate a list of criteria for poverty definition and a survey could include a check-box option to rank them - survey participants often like to have an "other" category where they can "fill in the blank" but that may not produce useful results, depending on the research question and proposed analysis. But the further issue on this is whether the data is USEFUL - we may determine a generalized definition of poverty but that is just the beginning of the next research issue - for instance, if we determine that "being hungry" is part of the definition, we probably need to go further to determine just what "hungry" means, how frequently "hungry" is experienced and maybe what the reasons for the hunger are. And once we have all the definitions pinned down, how will the same definition be applied to all loan applicants and is that an issue for Kiva itself, individual lenders or for the MFIs? And how will we know if each individual borrower meets our definition of "poverty" once it's established? Are we going to request a detailed credit report or continue accepting the recommendation of the MFIs that loans are worthy and the judgment of Kiva that each MFI uses "proper" criteria for their loans. These are all the types of issues that are relevant to creating a USEFUL study - otherwise, it's just promoting yet-another political agenda.

I would suggest finding a survey expert to help (and before anybody even thinks the thought, I'd love to have the time to do it but I'm way too overloaded to tackle a project like this)

Good luck with the survey and may we all continue in our quest for better lives!

	Author	Topic: Better polling needed (Read 3809 times)
0 Members and 5 Guests were last seen viewing this topic.