Hypothesis Identification Article Analysis

Hypothesis Identification Article Analysis

Research and Evaluation II

Gary Saxena

October 31, 2010

      I choose to research an article on cigarette smoking among gays, lesbians, and bisexuals because it seems that cigarette smoking seems to be a popular trend at their bars and night clubs. There is one “gay” bar in our town and it is often a lot of fun to go there to dance since they have great music there. Unfortunately, the cigarette smoke is almost too much.

      This article seeks to show evidence that the smoking rate among lesbians, gays, and bisexuals is higher than that of heterosexuals. Unlike the town I live in, the research was performed in a highly populated area with lots of gay bars to choose from. I live in a small Midwest town where being gay is not commonly accepted by all in the community.

      The data for this research was collected in association with the UCLA Center for Health Policy Research. The sample was taken by a telephone survey in which the numbers were randomly generated by a computer. The method of sampling used was stratified sampling in which the strata were counties in the area that were being sampled from. The proportions of participants in each strata were obtained from county population estimates. One possible flaw in this study could be that the respondents were asked to self-identify themselves as being homosexual. Due to the fact that some homosexuals were unwilling to tell the truth about their sexual orientation, the results of the study may be biased. Therefore, there may have been a vast number of “closeted” homosexuals in this heterosexual population. 

      This means that the analysis was of a logistic regression. This type of an approach is like that of the ordinary least squares regression, but instead here the responses are variable, rather than being normally distributed. Hence, linear regression would not have been a reasonable approach to this type of research. This research uses...