Fields such as human interaction are too unpredictable to be manipulated under laboratory settings. Human behavior, too, varies from place to place, and, therefore, results of one study that is conducted in one situation to one specific sample of people may not (and rarely do) have the same carryover effects to a different group of individuals conducted in a different spot or time. In other words, experimental, or quantitative, experiments necessitate that variables be closely matched and more, rather than less, homogenous. Therefore, they are best used in the hard sciences where they more often and most likely show consistent and repetitious results. The field of people is, however, more unpredictable and 'messy'. To that end, therefore, I find qualitative experiments more useful here. This is particularly so in this specific case where we seek to zone in on difficult patients and seek to understand their motives. The characteristics of 'difficult' patients vary from case to case and from time to time with, it seems to me, attendant impossibility of structuring any rigid, consistent rules. ' Difficulty' depends on patient, mood, context, and numerous other variables all of which differ according to individual case. This situation makes it particularly difficult to examine in a laboratory type of situation. The best that can be done, therefore, is to employ a qualitative approach that focuses on the individual person and uses rich interviewing in order to gain a better understanding of the situation.

The other problem...

Fieldwork, for instance, contains the advantage where the subjects can be mined (via observation or interviews for example) so that the subject can be explored in a deep way. The laboratory method, too, cannot always pin down metaphysical elements -- such as reasons for aggression (as in this case) - or human characteristics that vary and are contingent on specific situations or moods. Sometimes, too, cases can be too time-consuming to carry out, or certain groups, conditions or variables do not exist. It is for all these reasons that I found the qualitative studies useful in helping me form my decision regarding interventions.
The population of patients with dementia is not something that can be studied in the laboratory. It needs empathy on the research's part as well as recognition that these are people -- not subjects -- that are going to be studied. it, therefore, needs a lot of patience, creative ideas, and patient observation -- all of which is part of the qualitative approach. Each case, moreover, has to be adjudged on its own merit. Each case is different, each person is an individual, and with this in mind, an intervention can best be reached.

Sources

Creswell, J. (1998). Qualitative inquiry and research design: Choosing among five traditions. Thousand Oaks, California: Sage Publications.