Statistical analysis

We measured and compared rates of teaspoon loss over time using Kaplan-Meier survival analysis (Stata version 7). Rates of teaspoon loss were calculated using the formula:

L*100/D

where L is total teaspoons lost and D is teaspoon days (the sum of all the days on which each teaspoon was observed).

Study limitations

Our study design could have been improved by introducing control groups. The maintenance of an identical set of teaspoons in a locked but otherwise identical tearoom or cupboard could have been used to test for natural attrition. A placebo group consisting of non-functioning teaspoon replicas would have been a useful addition. Furthermore, the Hawthorne effect could have been a factor in this study; although discrete, the obvious red markings may have biased use of the teaspoons. It is possible that the teaspoons were devalued by the markings, and even though the act of teaspoon counting was carried out covertly after hours (so as not to waste precious work time), would-be teaspoon thieves may have been deterred from spoon lifting by the presence of a researcher.

The half life of teaspoons varied between the pilot and main studies. The relative endurance of the teaspoons in the main study may have been due to the higher total number of teaspoons—people may have been less likely to remove a teaspoon for personal use when plenty of teaspoons were available. It may be salient to determine what other factors can influence the half life of a teaspoon.