Breckler (1990) reviewed the personality and social psychology research literature and found several shortcomings of structural equation modeling, namely that model-t indices can be identical for a potentially large number of models, that assumptions of multivariate normality are required, that sample size affects results, and that cross-validation of models was infrequently addressed or mentioned. Many of the studies only reported a single model-t index. Breckler concluded that there was cause for concern in the reporting of structural equation modeling results. Raykov, Tomer, and Nesselroade (1991) proposed guidelines for reporting SEM results in the journal Psychology and Aging. Maxwell and Cole (1995) offered some general tips for writing methodological articles, and Hoyle and Panter (1995) published a chapter on reporting SEM research with an emphasis on describing the results and what model-t criteria to include.