Qualitative Data Analysis

I am learning so much in grad school, and arguably the skill in which I have improved the most (besides programming) is qualitative data analysis. I started using Excel back in the 6th grade for Science Fair projects. Crazy to think that was in 1999. I took advanced math courses throughout high school and then continued to use my analysis skills when working in psychology labs at Duke and Carnegie Mellon.

For this WIC project, however, there are no sophisticated quantitative analysis, at least that I have done so far, since almost all of my data is qualitative notes based on interviews.

The experience has been very enjoyable however and in a way has reminded me of detective work. I know the “culprit” (lack of awareness of children’s development) and am now looking for contextual clues as to how parents approach (or don’t) the problem. The focus of the interview is asking parents how they find information related to children’s health and why they prefer that method. I also present them with five different options for learning about children’s development (kiosks, booklets, texts, website, smartphone app) and ask them to rank them in order of preference. Again, I ask them why they ranked these tools the way they did.

While clearly each parent brings a unique perspective, three core values are consistently emphasized. Each one is also associated with a preference for learning about children’s development:

  1. Convenience/ease of access (prefer text messaging and/or phone apps)
  2. Interest/ desire to learn (prefer website and/or booklet)
  3. Trustworthiness of information (prefer kiosk in WIC/doctor’s clinic or asking a doctor in person)

It also contrasts interestingly with a text analysis assignment I did in an information visualization course I am taking. We had to find the “hidden terrorist plot” in a set of 50 FAKE (this was not real!!!!) documents using software our professor implemented. Here we had no direction however, so it was frustrating taking a shot at the dark. Additionally, in the scenario presented, we only had 50 documents and could not “look up” any additional. We could of course Google things to look for potential relationships between the entities mentioned in the documents (e.g., looking up where there are major airports in the US since several cities are mentioned in the documents as potential targets). For the WIC project, I had the benefit of interviewing additional parents, which I did, in order to get a clearer picture of patterns of behavior. I was not restricted in any way, besides the fact that I had to limit the duration of these interviews.


Spanish vs English interviews, gender roles in WIC

The WIC interviews have given me an interesting perspective as  a researcher due to my cultural background. While I have mostly lived in the United States, I lived in Medellin, Colombia from when I was 5-8 and speak Spanish at home. I learned both English and Spanish at the same time though my English is much better.

Many of the WIC clinics have a large Hispanic clientele ranging from 20-50+%. While there is no way to “prove” this, I hope that my pretty obvious ethnicity (due to my name and physical appearance) have made parents in general more at ease speaking to me. Besides the janitors, there isn’t a single man who works at the WIC! Add to that the fact that I am coming to the clinic as an outside researcher from Georgia Tech and it’s clear that there is potential bias of how the parents will perceive me.

Additionally, being able to speak Spanish has allowed me to speak with some parents who I would not otherwise be able to interview.

On a different but related note: I’ve read that there are slight personality changes when bilingual speakers use different languages. I’ve noticed that I’m slightly more upbeat yet also less talkative in Spanish. I’ve also been told that my voice is deeper when I speak in Spanish. This last point is subconscious because I never seem to notice, even though multiple people have pointed it out.

Interested related article on this last point:



iPhone vs Android design

Interesting points, though I disagree that the iPhone design was “designed by experts with superb taste” or that “the designers of Android, and companies like HTC, which modify that operating system in different ways, don’t have the same skills.” It’s weird because I made the same change and, while it may have something to do with the fact that I switched from iPhone 3GS to HTC one instead of the author’s iPhone 4->HTC switch, I immediately noticed a huge upgrade in the visual aesthetics, UI design, and interaction design. Sure, older Android phones are ugly as sin. This is majorly impacted by the fact that Android provides bland/hideous layouts for its developers whereas Apple’s defaults are very visually appealing. Android got the message though and has created more consistent (and most importantly, beautiful) themes for its phones. 



WIC Reflections – Marietta Interviews part 1

Research always takes about twice as long as you expect it to. Like in the Spring, the IRB has delayed things. I learned from the Spring to start early and had the Georgia Tech IRB by the time school started. This time, however, I’m dealing with double IRBs since WIC has its own. Complicating the matter is that each of the 200+ WIC districts has its own policy for dealing with research. Marietta, the last clinic I visited, gave me approval after 2 weeks. Still waiting on Lawrenceville (hopefully next week) and DeKalb (MIA). 

Yesterday I conducted two interviews at Marietta. They were originally meant to be practice to make sure there weren’t any flaws in the questions. I personally think they went well and only noted one thing to change in the interview. I present a figure with examples of formats for accessing developmental milestone information and include images of what they look like. I failed to include a picture of a kiosk, making it harder to imagine what it looks like.

Most importantly, both interviews took almost exactly 15 minutes. This study is challenging in that we need to interview parents while they are in the clinic since it would be difficult to schedule interviews in a different location. The WIC provides a secure and trusted location for conducting interviews, all critical features since this is already a disadvantaged and protected population. All the clinics have mentioned that families want to get in and out as quickly as possible so they can get back to work, a reasonable goal since sadly money is scarce.

After consulting with the nutritionists and directors, we decided the best time to conduct the interviews is right after the nutrition consultation. This is the second-to-last step involved and is followed by food vouchers. Usually it takes 15-20 minutes to print these, and sometimes slightly longer, so this is the most convenient time to do it.

For the two interviews so far, the nutritionist instructed a parent to come to me after explaining that the study was optional. This step is critical because, again, since we are working with a special population, we want to place even more emphasis on protecting personal rights. This is one of the reasons why the clause about the research study being optional is one of the first on the consent form.

Another interesting note: for the parents I spoke with, the nutritionists let me know that they “were very quiet and wouldn’t say much”. I actually found both to be fairly talkative. Hopefully this means I am doing a good job in seeming personable and not overly rigid? I will continue to consciously try and do that