2.5 Study design features
A study (Truswell et al. 1992) compared the health benefits of crisps ('potato chips') when fried in canola oil or palmolein. The article states (line breaks added):
Recently canola oil crisps have appeared in the market with the implication that their lower proportion of saturated fatty acids would have beneficial effects on plasma cholesterol.
We decided to examine this experimentally... a major manufacturer agreed to fry batches of potato crisps in canola oil or the conventional palmolein for a human trial. This provided an opportunity for us to run a double blind crossover trial.
Part of the protocol for the study is given below:
Subjects were asked to eat about a third of the day's crisps at the beginning of each meal and then, in the rest of their diets, to eat foods that were low to moderate in fat from a list of such foods...
The crisps were supplied in plain bags, unmarked except for two code characters: 'T4' or 'B9' in 1990 and 'D5' or 'GS' in 1991. Until the end of the experiment, the type of oil was known only by the food scientist who prepared the crisps.
In 1990, 21 subjects (12 men, 9 women) completed the 5 week experiment. They were randomly allocated to take palmolein ('B9') or canola ('T4') crisps for the first 3 weeks, then (without a washout period) changed over to the other type, canola or palmolein for another 2 weeks. Thus half the subjects ate (new) canola crisps for 3 weeks and the (usual) palmolein crisps for the next 2 weeks.
The other half of the subjects ate palmolein crisps for 3 weeks and canola crisps for the next 2 weeks. The first week of the 3 weeks served as an adjustment period. Fasting venous bloods were taken once at the start (usual diet), then for the last 3 mornings of both the first and second experimental periods.
- Explain how random allocation is used (and why), if at all.
- Explain how the impact of the Hawthorne effect has been minimised.
- Explain how the impact of the observer effect has been minimised.
- Would the carry-over effect play a role? Explain.