Thursday, May 08, 2014

Doctoring, researching, sampling

We've been having crazy weather for the past week. On Sunday and Monday, it was incredibly hot - rather unfortunate, as Monday was Memorial Day, and extra-hot weather isn't suitable for elderly people attending long ceremonies in graveyards. Tuesday was Independence Day; not so hot, but very cloudy and humid. At about three pm, winds started to blow and the humidity dispersed. Yesterday was ok - cool in the morning, warm in the afternoon - but in the evening, there was a tropical thunderstorm, and it's barely stopped raining since! This isn't Edinburgh with its four seasons in one day.

I've been very busy doctoring and researching during the past week. At the end of last week, I was working on the questionnaire, adding questions about spreadsheet competency and cognitive style. Some of the questions now have seven possible answers, so I've had to work on the exam program in order to display these possibilities. Moving backwards through the exam in order to allow for second thoughts is going to be interesting (in other words, hard to implement).

My mentor recommended for me a book called 'Business Research Methods' by Bryman and Bell. Its list price is 47 pounds although British Amazon is selling it for 38 (although they'll probably charge an arm and a leg for delivering it). American Amazon is selling it for $69, with the Kindle edition a mere snip at $61. There are over 800 pages in the latest edition, so it probably is value for money. 

Not unnaturally, with prices like these, I looked for the book on the Internet. Google books offers it, but I dislike reading a book from this site (it's okay for reading a few pages, but no more). I did find an ebook with the same title, written by a Dr Sue Greely, which I downloaded then Kindled the pdf. Unfortunately every page contained an advert and every chapter break contained several adverts, so reading the raw material on the Kindle was very difficult.

Thus I spent a few hours with Calibre: I converted the original pdf document to epub format, then edited this, removing all the adverts and cleaning up the text. Once I had a reasonably clean version, I copied it again to the Kindle. It turns out that this book is only about 80 pages long - it probably took me more time to edit than to read.

But there are some nuggets in the book. I learnt about the four kinds of variable: interval, ordinal, nominal and dichotomous. I discovered that my research mentions 19 (!) different variables, most of which are ordinal, some are nominal and a few are interval. 

The most interesting part of the book was about sampling. Obviously I am not going to research all the companies using Priority and I am not going to interview all the users within each company. So I need to use samples and I have to ensure that these samples are representative. I have a good idea how I am going to sample the users within a company, but I suspect that my sample of companies will not be particularly representative. This is partially because (until now) I have received no help from the company that develops Priority; they haven't even answered any of my emails.

[SO: 3577; 2, 13, 32
MPP: 396; 0, 1, 6]

No comments: