I thoroughly agree with Raffaele Zanoli in his assessment that the articles by Coimbra et al. (2020) and Vidal et al. (2020) are about as remote as can be from Q methodology, of which they claim that their procedure is a modification. What these many authors have done is to assemble a quite small sample of products –– N=9 whey dairy beverages in the one case and N=9 salted meats in the other –– and then instruct participants to assess these products using a Q sort on the basis of a relatively small number of descriptors (18 and 19, respectively). Factor analysis, however, was carried out on the beverage and meat products in relation to the descriptors, but not on the persons doing the assessing.

In one of his earlier papers, Stephenson (1936) showed that there were four basic systems of factor analysis

  • System I (R methodology), in which variables singly-centered around their own means were correlated across persons and then factored, resulting in a classification of variables.
  • System II (Q methodology), in which persons singly-centered around their own means (of M=0 in the Q sort) were correlated across traits or statements and then factored, resulting in a classification of persons.
  • System III (the transpose of System I), in which the variable scores are first standardized across persons and then transposed and restandardized across traits (double-centered), resulting in factors comprised of persons and thereby superficially similar to System II.
  • System IV (the transpose of System II), in which the trait scores are first standardized across traits and then transposed and restandardized across persons (double-centered), resulting in factors comprised of traits and thereby superficially similar to System I.

Only Systems I and II were regarded as truly of methodological proportions –– I dealing with what is objective in the world and II with what is subjective; III and IV were only logical possibilities and mere applications of the more fundamental systems. Taking repeated measures across different occasions, sometimes referred to as O technique (Burt, 1981; Cattell, 1951), was also simply an application lacking methodological status (Stephenson, 1952).

From this vantagepoint, what Coimbra et al. (2020) and Vidal et al. (2020) have produced is something akin to a System IV analysis rather than a modification of System II. Their focus is on variables rather than on the subjectivity of the persons doing the assessing and who may have been blissfully ignorant of and indeed unconcerned with the variables of interest to Coimbra, Vidal, and their associates, who for their part have no obvious interest in what the participants were thinking. It is striking that the authors cite only one or two authors who have written about or utilized Q methodology.

In tracing the point at which Coimbra and Vidal lost the scent, it would be important to examine the article by Brard and Lêa (2018), whom they cite and claim as their authority. Brard and Lêa give acknowledgement to Marcel Trudel and cite texts by Brown and by Watts and Stenner, thereby testifying to at least rudimentary knowledge of Q methodology. As Coimbra and Vidal indicate, Brard and Lêa proceeded in much the same fashion and pursued much the same goals with their sample of 39 shampoo perfumes, but at least Brard and Lêa recognized how Q methodology was different in intent from their own procedures and goals, even while regarding what they were doing as in some way a modification of Q methodology. How different their study would have been had they been interested in their participants’ thoughts and feelings about perfumes rather than in the characteristics of the perfumes themselves; i.e., in traits rather than persons instead of the other way around (Stephenson, 1935).

References

Brard, M., & Lêa, S. (2018). Adaptation of the Q-methodology for the characterization of a complex concept through a set of products: From the collection of the data to their analysis. Food Quality and Preference, 67, 77–86.

Burt, C. (1981). O, P. Q, and R techniques. Operant Subjectivity, 4, 102-119.

Cattell, R.B. (1951). On the disuse and misuse of P, Q, Qs and O techniques in clinical psychology. Journal of Clinical Psychology, 7, 203-214.

Coimbra, L.O., Vidal, V.A.S., Silva, R., Rocha, R.S., Guimarães, J.T., Balthazar, C.F., Pimentel, T.C., Silva, M.C., Granato, D., Freitas, M.Q., Pollonio, M.A.R., Esmerino, E.A., & Cruz, A.G. (2020). Are ohmic heating-treated whey dairy beverages an innovation? Insights of the Q methodology. LWT – Food Science and Technology, 134, art. 110052.

Stephenson, W. (1935). Correlating persons instead of tests. Character and Personality, 4, 17-24.

Stephenson, W. (1936). The foundations of psychometry: Four factor systems. Psychometrika, 1, 195-209.

Stephenson, W. (1952). A note on Professor R.B. Cattell’s methodological adumbrations. Journal of Clinical Psychology, 8, 206-207.

Vidal, V.A.S., Paglarini, C.S., Freitas, M.Q., Coimbra, L.O., Esmerino, E.A., Pollonio, M.A.R., & Cruz, A.G. (2020). Q methodology: An interesting strategy for concept profile and sensory description of low sodium salted meat. Meat Science, 161, art. 108000.

Leave a Reply