Study Overview
Catherine Crisp reports in the article “Correlates of Homophobia and Use of Gay Affirmative Practice Among Social Workers” that research on social work practice with gay and lesbian clients has historically focused on examining attitudes towards individuals. Prior to the reported research, no studies have researched the relationship between attitudes and practice using validated measures. The innovative Gay Affirmative Practice Scale (GAP) is used to examine the relationship between attitudes and practice. The study focuses on correlates of homophobia in social workers, correlates of a measure of GAP in social workers, and the relationship between homophobia and GAP in social workers. The study’s strengths are the large sample size, random selection from the global largest social work organization, and the focus on direct practitioners. The obvious limitation of the study is the extremely low response rate.
Design Overview and Critique
Because the study focused on the link between practice and attitudes with gay and lesbian clients, only social workers who provided the direct practice to clients were included. 1,500 NASW members were planned to be surveyed. The survey package was mail to each member. The package consisted of four elements: GAP and Attitudes Towards Lesbian and Gay Men Scale (ATLG), questions on respondents’ demographics, a cover letter with informed consent information, and an envelope to return the questionnaire. The study is descriptive.
The chosen study design is strong enough to gather reliable information. The questionnaires are targeted at practitioners in the field of research. The survey package includes all necessary information. However, the direct mailing contributes to the low response rate due to the absence of confidentiality. In addition, the length of the questionnaire is not motivating – respondents may be reluctant to spend a lot of time proving the answers. Another weakness is related to the limitation of response time (two weeks). The chosen study design (direct mailing) is not the most effective tool to gather sensitive information. From one side, it is rather cheap. On the other side, the response rate is extremely low. Thus, unnecessary biases are created.
Survey Tool Critique
The survey is carried out with the help of a questionnaire. The questionnaire is sent to randomly chosen respondents. The weakness of the survey tool includes a low response rate caused by a lack of control over delivery. Moreover, the failure to guarantee confidentiality results in decreased motivation to reply to questions. The results reveal that the majority of respondents are married, heterosexual females. Thus, the results of the study are questionable because the attitude of people with different demographic characteristics is not assessed.
Sample Overview and Critique
The sample consists of social workers who provide direct practice to clients. 1,500 NASW members whose function was self-identified as a direct practice were randomly selected from the membership roster.
The sample size is the key strength of the study. The respondents are directly involved in the practice under research. Moreover, random sampling contributes to the reliability of gathered information. Random sampling in the given study helps avoid gender biases. Nevertheless, the weakness of the sampling is the uncontrollable response rate. The mailing technique does not guarantee adequate reach to potential respondents. Low compliance does not reflect the general trend. Selection bias is not present. Moreover, the given study is descriptive and it requires hundred of subjects to give acceptable confidence intervals. Overall, the sample size is enough to guarantee the validity
Independent and Dependent Variables
Independent variables are demographic characteristics of respondents: gender, race, and sexual orientation, religious and political affiliation. Dependent variables are training on gay and lesbian issues, feelings about and contact with gays and lesbians, and social desirability. The dependent variables are correlated with the independent variables. Thus, there is a direct correlation between sexual orientation and feelings about gays and lesbians.
Instrumentation Overview and Critique
Three measures are used: the HATH, the ATLG, and the GAP. The HATH scale assesses cognitive beliefs about gays and lesbians with a five-point Likert-scale response rate. The HATH is validated with college students (reliability of.86) and the Brown prophecy formula (reliability of.92). The ATLG scale assesses affective responses towards gay men and lesbians and has two subscales to assess attitudes towards lesbians and attitudes towards gay men as separate constructs. ATLG has been validated with members of gay and lesbian organizations demonstrating high internal consistency with an alpha of.90. Finally, the GAP is a two-dimensional instrument developed to assess the degree to which practitioners hold beliefs (initial reliability is.95; the Standard Error of Measurement 1.91 for belief domain and 2.71 for behavior domain).
All of the above instruments have been validated and proved to be effective in the assessment of attitudes towards gay men and lesbians. The Likert five-point response rate is traditionally effective in survey-based studies. The relevancy of ATLG and GAP instruments contributes to the high validity and reliability of results. Moreover, the validity for each instrument is provided. Moreover, each of these instruments has already been tested with college students. Internal consistency is high. There are no weaknesses of these instruments because they are highly relevant to the study objectives.
Validity Critique
Validity is positive because the research truly measures the attitude towards gay men and lesbians. The means of measurement are accurate and they are measuring what they are intended to measure. In particular, the HATH scale measures the cognitive beliefs about gays and lesbians. The validity of the study is further strengthened with the ATLG scale due to two subscales to assess attitudes towards lesbians and attitudes towards gay men as separate constructs. The ATLG scale helps to avoid gender bias. Moreover, the ATLG has been validated with members of gay and lesbian organizations. The internal consistency is high. The final instrument, the GAP, contributes to positive validity as well because it is used to assess the degree to which practitioners hold beliefs. In summary, the validity of the study is high enough to guarantee the reliability of the results.
While the validity of all instruments is justified (relevance to the field of research is clearly stated) and proved (all relevant tests are made), additional measures could increase the validity as well as consistency of the research findings. Moreover, the information presented on validity is scarce and hard to identify. Only a couple of phrases are written on validity. Therefore, additional comments would significantly contribute to the validity of the presented research.
Reliability Critique
The study reliability exceeds.9 as reported in the methodology section. Pearson’s r was computed to identify the relationship between internal-level variables and scored on the HATH, ATLG, and GAP scores. The relationship between measures of homophobia and gay affirmative practices is traced. The relationship between nominal variables and scores on scales was tested with T-Tests. In addition, a one-way analysis of variance was used.
Research instrument repeatability and internal consistency are proved. Therefore, reliability is proved as well. The instruments are valid due to the previous testing on similar samples. Research results can be applied to a wider group. Because of study focus on a very specific social issue, similar results would be obtained if another group containing different respondents were used. In addition, the method of data gathering leads to consistent results. Replication of the study can produce similar results. Nevertheless, there is a number of factors contributing to the decrease in reliability. In particular, low response rate and direct mailing survey tool do not reflect the general tendency. Thus, the results may be reliable and should be further tested. Moreover, the instrumentation and design choice are not justified in the methodology section. The additional instrument could generate more reliable findings.