Comment – 1. Sample size (Donald F. Klein)

Katz  states, ”In this book I describe the research approach, and the new findings that led to: (1) identifying the major  mood, cognitive, and behavioral components of the multifaceted depressed state; (2) Uncovering the dimensional structure of the disorder; (3) Further elaboration of the psychological turmoil that defines the experiential  state of depression: (4) Proposing a new theory about its conflictual nature detailing the interaction of neurochemistry and behavior which comprise the state, and (5) Describing the impact of the antidepressant (AD) drugs on behavior and chemistry, that is, the drug- specific actions on behavior, and the onset and sequence of clinical actions that precede recovery". (p. viii).

Katz believes that adequate description of depression requires contributions from doctor observations, patient self-reports, psychomotor performance, nurse observations and video interviews coded by behavioral evaluation scales. (p 26-8). These observations are linked, in part, by factor analyses. .

Katz’s “constructs of the depressive disorder are based partly on phenomenological analyses from Grinker et al [4]  and Kendell [7], and partly, on the result of factorial analyses of data assembled from the one hundred four  moderately to severely ill patients sampled across the six hospitals in the CDS [1].The constructs encompass affect or emotional components such as depressed mood, anxiety and anger, disturbed psychomotor performance, thinking , somatic functioning and social behavior elements.  There are 11 components inter-correlated in various degrees that were factor analyzed to derive fewer dimensions, independent  in quality that could be applied to understanding the structure of the psychopathology underlying this class of disorder” (Katz p.26).  

My general concern is that the sample sizes, a total of 106 patients, derived from six sites, are very small to serve as the bases for stable, generalizable factors. Further the sample sizes seem to fluctuate. For instance on p.29 the sample is stated as 130.

Do you believe that this sample size is adequate for your purposes? 


Grinker R, Miller J, Sabshin M, Nunn R, Nunnelly JC. The Phenomena of Depression. New York: Hoeber; 1961.

Katz MM, Koslow S, Berman M, Secunda S, Maas J, Casper R, Kocsis J, Stokes P. Multivantaged approach in the measurement of behavioral and affect states for clinical and psychobiological research. Psychological Reports 1985; 55: 619-91.    

Kendell RE. The Classification of Depressive Illnesses. London: Oxford University Press; 1968.


Donald F. Klein

January 30, 2014