In the social sciences, the two main empirical research traditions are the quantitative approach that draws on techniques such as experiments, surveys, histories, analysis of archival information, etc., and the qualitative one that utilizes techniques such as case studies, participant observation, open interviews, etc. (Frankfort-Nachmias and Nachmias 1996:281-6, 304-9, Creswell 1994: 10-12, Patton 1990:14, 36-9, 162-6, Yin 1989:8, 17-22, Day and Castleberry 1986:94-5, Vinehall 1979:108-15, Bogdan and Taylor 1975:3-7, LaPiere 1934:237). The labels 'quantitative' and 'qualitative' research are interchangeable with notions like 'objective' and 'subjective' research (Burrell and Morgan 1979:1-8).1 The quantitative approach tends to be related to logical positivism, the traditional empirical research paradigm of the natural sciences, while the idea of qualitative research relates to a phenomenological, hermeneutic research tradition that originated in the social sciences. The phenomenological approach can be traced back to the sociologists Max Weber and George Herbert Mead and the philosophers Edmund Husserl and Ludwig von Wittgenstein.
Both the quantitative and the qualitative approaches have strengths and weaknesses. Their respective strengths and weaknesses can be discussed along three dimensions: external validity, reliability and precision. External validity (or 'ecological validity') refers to the possibility of generalizing findings on a researched population to other populations, ideally to populations as they can be found in the 'real world'. External validity touches upon issues related to the relevance of research (Coolican 1990:36, Yin 1989:41-
4). Reliability refers to the possibility that a study can be repeated by another researcher and yield the same results (Silverman 1993:145, Coolican 1990:34, Yin 1989:41, 45).2 Precision refers to the numerical accuracy that can be attributed to research findings. It relates to the quantitative significance of empirical research (Yin 1989:40-3, Weick 1979:35-6, Thorngate 1976:406).
Quantitative and qualitative research acknowledge different trade-offs regarding external validity, reliability and precision (Stoecker 1991:92-3, Coolican 1990:34-8, Patton 1990:14, Bruner 1979:2-3, Weick 1979:35-6, Thorngate 1976:406, LaPiere 1934:237). As LaPiere pointed out in the 1930s:
Quantitative measurements are quantitatively accurate; qualitative evaluations are always subject to the errors of human judgement. Yet it would seem far more worth while to make a shrewd guess regarding that which is essential than to accurately measure that which is likely to prove quite irrelevant.
Quantitative methods, through their strict control and manipulation of their research environment, e.g. a laboratory, tend to have an edge over qualitative methods with regard to precision and reliability. The possibly lower precision of qualitative methods is due to their information richness (Stoecker 1991:91, Cohen 1989:13, Bartlett 1932:12). For instance, in qualitative research precision cannot be expressed through numerically calculated significance levels.
A quantitative approach may have weaknesses with respect to external validity. The rigidity of control and the artificiality of the research environment— normally the laboratory—pose threats to external validity. Irrelevance, triviality and a weak record regarding external validity are evident in much quantitative research in cognitive psychology. Such stern criticism comes both from within the field (Varela et al. 1991:xvi, Coolican 1990:36, Gardner 1987:134, Tulving 1985:395, Bryce 1985:79, Bruner 1979:28, Neisser 1978:3, 11-12, Bartlett 1932:2-9; see also chapter 2), and from a qualitatively committed research community (Silverman 1985:43, Harre 1981:4, 15-17, Gross 1974: 42, Rowan 1974:86-7, 96, Husserl 1954:201-12).
In short, the results of a hundred years of the psychological study of memory are somewhat discouraging. We have established firm empirical generalizations, but most of them are so obvious that every ten-year-old knows them anyway.. We have an intellectually impressive group of theories, but history offers little confidence that they will provide any meaningful insight into natural behaviour.
Similar accusations have been voiced for consumer behaviour research (Shimp 1994:1-4, Wells 1993:4903, Foxall 1993:46) and for applied cognitive psychological research (Herrmann and Gruneberg 1993:553-6, 562). Methodological pluralism has been called for in consumer behaviour research to amend that situation (Foxall 1993:46) but little has been achieved in that respect so far. The accusation that most consumer behaviour research has been motivated by 'the availability of easy-to-use measuring instruments.and/or the almost toy-like nature of sophisticated quantitative techniques' (Jacoby 1976a:2) has been upheld over the past two decades since first voiced in the mid-1970s (Shimp 1994:5, Wells 1993:490; see also Jacoby 1976b, LaPiere 1934:237). When it came to the structuring of scientific research, apparently, empirical research has driven conceptual research rather than the other way round.
The main reason for the lack of external validity in quantitative research is the artificiality of the laboratory environment which requires the researcher to isolate and reduce 'units of research' in order to make them quantifiable and controllable. Often 'non-meaningful units of behaviour'—taken out of their natural context, and sharply isolated from other units—are researched. Bruner (1979:28) notes: 'The more rigorously isolated from context and the more tightly controlled the conditions of experiments, the more precise and the more modest have results been.' In addition, convenience sampling that is based on untypical populations, such as undergraduates, is frequently relied upon in quantitative psychological research,3 which poses a further external validity problem. In general, an external validity problem is not an insignificant problem since 'laboratory research is not meant to be generalized only to other laboratories' (Kidder 1981b:252, also Wells 1993:492).
External validity is likely to be higher in qualitative research since the context and the phenomenon under investigation are not artificially separated (Coolican 1990:237, Cohen 1989:13, LaPiere 1934:236-7). Meaningful units of research are approached on a more comprehensive basis as they are likely to occur in 'real life'. There has been a strong call for so-called real-life studies of psychological phenomena from within the field of psychology (Alba et al. 1991:36, Gardner 1987:135, 258, 298, Neisser 1978:11-14, Thorngate 1976: 406-7, Neisser 1967:305, Bartlett 1932:12). Qualitative research, despite being rare, has a long tradition in psychology and it has yielded valuable contributions in cognitive and in social psychology, for instance Sacks (1991), Luria (1987), Sacks (1986), Janis (1983) and Bartlett (1932) who also refers to Cassel (1895); and also Lazarsfeld's studies in consumer psychology in the 1930s (Fullerton 1990, Neurath 1988). However, as indicated, real-life studies come at a methodological cost regarding precision and reliability. Certain techniques may be applied to increase the precision and the reliability of qualitative approaches, but even then a quantitative approach is likely to be superior in that respect.
Was this article helpful?