This is an interesting question, and we can look at it using the bell curve,

In a bell curve, you have the most activity at the top of the curve, which can be translated into the mean. The further you deviate away from the mean, the less and less activity you see. In fact, if you go 1 standard deviation away from the mean, you encompass 64% of the population. 2 standard deviations from the mean would encompass 97% of the population. 3 standard deviations from the mean would encompass 99% of the population.

The bell curve is based on the average activities of many people. In a sample you hypothetically need 30 individuals to be included in the mean in order to make sound statistical inferences.

So let's break it down. A representative sample would include the activities of at least 30 people. Lest look at food shopping for example. Some people shop 2 times a week, while others shop 7 days a week. The numbers average out to 3 times a week.

If you examined the person who shops 2 times a week, they are below the average and are not representative of the sample as a whole. Neither is the person who shops 7 days a week.

Therefore, the answer to the question is simple. No, you can't generalize one person's activities for the rest of the population. The only way you can understand the population is to take a random sample of at least 30 people and analyze their activities as a group in order to make conclusions about the population.