Burkard, Alan W. And Knox,

Burkard, Alan W. And Knox, Sarah (2004). Effect of therapist color blindness on empathy and attributions in cross-cultural counseling. Journal of Counseling Psychology, 51:4, 387-397.

Whether research investigations are designed on the basis of an experimental, descriptive, historical, or case study format there must exist a clearly defined research question which informs the reader as to the specific intent of the research undertaken. When the research design is experimental or descriptive there must follow a well-formulated testable null hypothesis. Should these two research tenets not be in place then the research is said to be ad hoc and of little useful propose with respect to the validity and reliability of content knowledge. The present article reviewed and authored by Burkard and Knox (2004) can be best described as ad hoc research and did not follow the best-fit practice for effective research. The reasons for this summary statement is based upon the following reasons:

research question is to be presented at the beginning of a research investigation and if designed as an experimental of descriptive study a testable null hypothesis (es) follows. Buckard and Knox did not formally present a research question and the hypotheses presented were not only directional and non-testable (i.e., null form) but also highly biased as well: "...we hypothesized that those therapists were higher in color blindness racial attitudes..." "...we predicted that therapists' color blindness would be related to...." "..we predicted therapists' color blind racial attitudes would..." (p. 389). Continuing one step further, the research investigators failed to even alert the reader as to the research problem even though they presented a great deal of information pertaining to the overall area of study. At no time is a research investigator permitted to pre-formulate a direction that a sequence of events might take. All that a research is permitted to do is hypothesize that no differences, effects, or relationships will be uncovered as a result of the investigation. The reason being is that without a testable null hypothesis a research investigator cannot explain why something some thing did happen, and not why something did not happen. Therefore a research investigator must always begin with a statement (null hypothesis) that something will not happen.

With respect to the sample used in the study the authors failed to alert the reader as to the randomness of selection. When samples are chosen on the basis of non-randomization, the statistical tool chosen to analyze the day must be "re-fitted" to accommodate for non-randomization of selection.

The data analysis discussed in the report (p. 391-393) makes reference to an ANOVA and Pearson "r" coefficient analysis. These two particular statistical tools are employed to determine effects/differences and relationships between and amongst the variables under investigation. As there exists no rationale for seeking differences or effects the use of the ANOVA technique is circumspect. For there to exist legitimacy for the use of the ANOVA technique the reader should have been presented a testable null hypothesis stating effects and/or differences between and amongst the variables identified. In addition the research investigators did not identify any nested and/or intervening variables that might possibly impact upon their study. Such variables might have been in the form of gender, time in practice, practice demographics, practitioner ethnicity/race, and socio-economic status of client.

Other mishaps with respect to the reviewed study include the lack of identifying the independent variables (even tough the authors make mention of such variables but without explanation and definition on page 391); an adequate description of the IRI and MCSDS validity coefficients; a rationale for investigating a culturally sensitive area with measurement instruments that are over 20 and 40 years old; and drawing conclusions on the basis of non-empirical investigation.

Although the article provided interesting reading this reviewer cannot recommend the information reported to be embodied within a larger content knowledge area with respect to the understanding of cultural diversity

2. Vazire, Simine and Gosling, Samuel D. (2004). e-Perceptions: Personality Impressions based on personal websites. Journal of Personality and Social Psychology, 87:1, 123-132.

Any research investigation must command enthusiasm, interest and passionate commitment. If these tenets are not adhered to then the research is merely an ad hoc exercise and not a scientific endeavor of importance. Further, research investigative situations that make use of numerical data must strictly adhere to a very stringent format that includes, but not limited to, a proposed research need, a stated research question, a formulation of a testable null hypothesis, a definition of the variables under investigation, and a selection on an instrument that will produce the needed data for statistical analysis. The article published by Vazire and Gosling will be evaluated with respect to these requirements.

Although the authors do define "personality impressions" (p 124) they fail to present the reader with an acceptable reference for definition. In other words, there exists no supportive information with respect to the legitimacy of their definition.

Even though an evaluation of definition is not an issue that comes first in a critical evaluation of a research endeavor, in the present situation it is one that unfolds in front of the reader immediately.

The authors abided by the requirement to state a research question in the beginning of the investigation (p 124); however, they failed, like Burkard and Knox, to present a testable null hypothesis. As mentioned in the evaluation of the first article, whenever statistical tools are employed in a research investigation a null hypothesis must be presented for statistical testing.

With respect to the personal websites contact to be included in the study the authors failed to inform the reader as to the total amount of possible websites and the method used to identify such websites. Further, the authors failed to inform the reader as to the demographics of the websites as well as to their classification. Simply by stating that they are personal websites without classification is unacceptable in terms of sampling procedure. As a side bar note one would have to ask the authors whether or not the 385 authored websites were business related, sex related, issue related, or even profession related. These intervening and nested variables will, and can, impact the results of a study tremendously.