After work, I came home and turned on the TV to VH1 Europe. I’m pretty sure only me and Lanz Stopar of Slovenia were watching, because he was the only person who requested things on VH1 Jukebox. One of these songs was Cher’s “If I Could Turn Back Time.” Upon further review, I’d like to amend my statement that she was wearing assless pants in this video. While I have a particular pair of assless pants in mind, she was not wearing them in this video. Instead she was wearing a black strip of fabric that made a V in the front covering her hoohah and boobs, and was basically a thong in the back, She had some kind of fishnet body suit over it. Amazing.
This video was so amazing that it killed the power. This was actually the second time power died where I was yesterday. When I was at the grocery store that morning the power went out briefly. I’m glad I got that flashlight. I put up the bednet and watched Dexter. The power came on after a couple of hours, and I went to bed because I got so sleepy in the dark.
I woke up at quarter to five, reaching new heights of ridiculousness in my sleep schedule. I had some peanut butter and bread (Malawi peanut butter is really tasty, by the way), and then bathed. The shower/bath went much better today, as I was more efficient about washing my hair first while the water was definitely hot, and then putting water in the bathtub for the rest of washing. So I was warm the whole time. I wandered out around 6 to see if the supermarket was open, but it doesn’t open until 8. Maybe I’ll try to go tonight before it closes at 6, or just wait until the weekend. When I returned, the man at the Villa who seems to be in charge of things reminded me that they serve breakfast, with the implied message that I should really come down and eat. So I did after a while, and they fed me tea and toast. I figured this would be a cup of tea and a plate of toast, but it was much more elaborate. First, there was a tray with a teacup and saucer, the teabag, two spoons, a pitcher of milk, and a whole bowl of sugar. Then there was another tray with a plate with two slices of toast, another plate with two pats of butter, and an opened can of the strangest jam I’ve ever had. It was like spreadable candy. But it had a taste I acquired once I realized the butter was kind of odd. There are a lot of dairy products around here that apparently don’t require refrigeration, and they are interesting.
I came into work and called the district coordinators. One said he would email me the information I needed, but I got a message later that he had called to say the power was out there. The other talked so fast that I could not understand him, but I think he told me that he needed to get one last piece of info, and it would take 2-3 days. After that I tried to talk to Joe again on Skype, with hilarious results. The delay is just amazing. The chat is much faster, so he’d type something, and then about 10 seconds later I’d hear him type. It also apparently hogs the limited bandwidth we have here… it’s like using the internet in 1996!
Kirsten and I went to a meeting with the guy in charge of the consulting company running the survey. I came to realize how difficult sampling really is. It sounds so straightforward in my proposal: “In each of the four districts, 240 respondents age 16 to 65 will be recruited, for a total of 960. Multi-stage sampling will be used to select respondents. Each study location is divided into a number of enumeration areas (EAs). Through proportional random sampling, EAs will be selected from each study location. Within each EA, a list of households will be created and households will be sampled in proportion to the number of them in the EA. Within each sampled household, the interviewer will make a list of all household members 16 to 65 years old and then will randomly select one to be the respondent. Either one young adult (younger than 24 years old) or one older adult (24 years or older) will be selected from each household on a rotational basis; this is to ensure that both younger and older adults are sampled in adequate numbers for comparison purposes.”
But this does not really being to describe what is going on. Within each district there are subdistricts called Traditional Authorities. Within each TA there are villages. We need to know the populations of these so that we can randomly select a proportional number of people to be interviewed. Getting this information is difficult enough, apart from its accuracy. All this sampling ahead of time tells you is how many houses per village to visit. It doesn’t nail down who exactly to visit. This happens in the village during the field work. Basically, you have all these interviewers who scatter out to places, and somehow at the end of the day you want to get not only the right number of interviews, but you also want half male, half female and half youth (under age 25) and half adults (overall and by gender). And the field teams have to figure out how to get from place to place with maps that may not be accurate and that don’t show you important information such as, it’s not actually a mile to the next village, because there’s a huge hill in the way.
So, there’s that. I don’t envy the folks who have to figure that out. In the course of discussing the survey (which I have to go through and renumber, because it’s a mess, and it will be a bitch to try to keep question numbers the same from the last midterm when there are new things inserted everywhere), we came upon a major question. There are 8 districts in which BRIDGE has been doing the project. We surveyed 4 in the midterm survey, with the intent to survey all of them in the final survey. But USAID added 2 years to the contract, so the final survey will be in 2009. Hey, we said, why don’t we do a midterm survey in the other 4 districts. So now, a year and a half has passed between the last midterm survey. The question was raised, how do we compare the midterm survey results, because not only are they completely different districts, they are completely different time points. Which led Kirstin to ask, very astutely, does it make sense to do this survey in the other 4 districts? Does it make more sense scientifically to re-do the first four, so we can compare over time? This is a really good question, so we’re going to talk to Rajiv today to figure out whether we need to hurry and change plans while we have the chance.
We came back here, where I was almost done typing this when the power went out. Kirsten said they’re cleaning the turbines at the main power station, which has reduced their capacity, and thus there have been rolling blackouts. This particular one killed my computer in a way that caused me mild panic for several minutes. Once I removed and replaced the battery, all was well, and now I’m plugged into the power strip instead of the wall in the hopes that this will do something better should the power go out again.
I think I’m going to wander out for some lunch… maybe today I will try some local food!
