This chapter focuses on the need for order in running an experiment. It begins by addressing the need for having a specific form to collect the data and then proceeds to address some of the issues that every experimenter should be aware of when doing surveys and experiments.
Before you accumulate any data, you must design the questionnaire (for a survey) or have a very good idea about your operational definition and process (for an experiment), and you must have an appropriate form for collecting the data. It sounds both simple and silly to talk about these items, but they are fundamental to the integrity of the study.
For example, instead of asking directly for a person's age, you may want to request the actual month, day, and year of birth. Some people, especially as they grow older, have unpredictable systems for altering their ages at each birthday. Asking for the date of birth, instead of asking directly for age, increases the likelihood of an accurate answer. The computer will calculate the exact age later. Never ask interviewers or respondents to calculate anything themselves . Get the raw numbers , and let the machine do the arithmetic. If you are interested in the ratio of a person's weight to height, for example, the interviewer or respondent should record the weight and height, and you should leave the ratio to the computer. This procedure saves time and increases the accuracy of the results. Computers divide better than distracted people with calculators do.