Making a factor in an experiment measurable depends very much on whether the factor is being treated in a qualitative or quantitative manner.
Qualitative measurements are often described as being "opinion-based", but it might be better to describe them as measurements that aren't attached to an empirical value. In contrast, a quantitative measurement is an empirical value such as the length or weight of an object. Qualitative measurements can include numbers, such as asking medical patients to rate their pain level on a scale of 1 to 10, but these numbers are somewhat "open to interpretation" rather than empirical, verifiable levels or amounts.
Some of the responsibilities in thinking and argument for the investigator are;
- to decide whether qualitative, quantitative, or both methods are the most appropriate for a particular factor, and explain how those methods are being implemented,
- to acknowledge and accommodate the limitations of their chosen approach and implementation,
- to choose an appropriate sample size, and represent the results in a manner that recognizes the relative impact of the data
For example, in my teaching career I've often encountered problems with the way that attendance data is analyzed.
In reference to point 1, attendance is predominantly a quantitative factor, but we frequently don't know what the minimum requirement is for a student to be counted as "present". In some schools, even a single "tardy" attendance mark causes that student to fall into a different attendance category; at other schools, a student can be tardy or absent for virtually the entire day, but as long as they are present for even a few minutes, they are counted as present for that day. This emphasizes the responsibility of the investigator to explain the terms of the values being acquired.
In reference to point 2, it is difficult to distinguish student absences that are due to different factors, such as sports, juvenile delinquency, family issues, medical appointments, and clerical error. Each of these has significantly different implications for the means of addressing the data, but because they all simply appear in the same tally, "absent", it becomes difficult to determine an appropriate method of addressing the issue.
In reference to point 3, at one point we were shocked to see that a certain student demographic group was absent from school on more than 40% of school days; this was more than triple the percentage of the next closest group. It took a considerable amount of investigation to learn that this particular student demographic group was composed of only two students from the same family. The fact that the sample size was two students, out of over 2500 enrolled in the school, should have marked this group as an outlier and therefore of considerably less concern regarding our overall outcomes, which emphasizes the responsibility of the investigator to understand and represent the data in an appropriate context.