So, economists think consumer make their purchases rationally, and that they make their decisions based on past behaviour and expectations of the future.

I put that claim to the test, with the sample size of 1 - me. For the past 20 weeks or so, I’ve been making grocery runs every Saturday. I only do my groceries once a week, since I am fairly capable of gauging my expectations with regards to what I want to eat for the week.

I base my diet on staples - there is a set amount of carbs, proteins and veges I eat each week. Because the amount of veges I eat is embarassingly dwindling (I eat a lot during the beginning of the semester, and less and less as time goes on), and my protein intake is relatively fluctuant to my desires at the time of purchase (i.e. if I feel like eating a meat on Saturday, I will buy meat for the rest of the week, even though there will be some days I totally do not like to eat meat). So, the only data I analyzed is my weekly carbohydrate purchases.

There are primarily 4 forms of carbs that I buy - bread, Indomie, instant noodles and pasta¹. I also buy and eat plenty of rice, but rice is unquantifiable (it actually is, but rice isn’t as instant as eating bread or instant noodles). So, how did I fare?

Pretty horrible graph isn’t it? Not only it tells nothing about my spending habits, it also looks bad to begin with. So, let’s look at some numbers instead.

As it turns out, due to the layout of Coles, I usually buy my bread first, followed by Indomie and instant noodles, and then spaghetti (or whatever pasta that tickles my fancy). So, by virtue of backwards induction, we can see that my bread buying behaviour is affected by my Indomie, instant noodles and pasta buying behaviour. Ergo, when you regress buying behaviour of the rest to bread, you get this:

There you go. Indomie and bread seems to be inversely proportionate with one another (and is significant about the 90% level too). Hence it can be said here that, if I am expecting to eat a lot of Indomie, I buy less bread and pasta. But what about past behaviours? How would past behaviour affect current behaviour? Here’s a regression with a 1 period lag:

From the table above, my bread buying behaviour can be seen to be quite affected by the previous week’s purchases. The more bread I bought the previous week, the less I’ll buy this week. Same goes with noodles. Too bad none of them is really significant (except Indomie and the previous week’s amount of bread I bought).

Now, this is not inaccurate at all. One can get sick of staples. Also, I buy ahead for dry staples like Indomie (I typically keep about 2 bags of Indomie at home, just so that in case I get hungry I get something to eat). So, from what the data suggest, I would buy bread, get sick of bread, replace with Indomie, get sick of Indomie… ad nauseum.

I noticed some interesting trends. When it is close to exam time, my carb intake increases a lot. Hmm.. this is why I put on weight I suppose LOL.

And yes, this is an escape from studying. Oh well, back to hit the books.

1. bread in loaves, Indomie in bags - one bag has 5 packets, instant noodles in packets, pasta in packets [↩]

Cool? DiggDel.icio.usTechnoratiFeedsterFurl
RSS feed for comments on this post  |  TrackBack URI for this post