Population, sampling frame, unit of analysis, and sample types

¶ … studied is the residents of the geriatric facility (1161). The sampling frame is whether the patient fell in the previous year or not. The unit of analysis is individuals. The type of sample is random. This strategy is sound for this type of study. The population is divided into two groups, one being a control group. This will allow the researcher to test for the different medical factors, and see the differences between the two sets. For studies like this, randomization is the key to the study being free from bias. When the study is fully randomized, then it can be said to be free from bias. Additionally, this study makes note that its population is just the residents of this particular home. If the study was about all elderly people, then choosing only the ones from this home would not be randomized, and the results would be subject to more bias. However, for this study, the method of sampling is good.

b. The population is the rental housing stock of the city. The sample frame in this study is the different ages of the buildings. The study will result in vacancy rates for each age category of building, and then an overall vacancy rate. The unit of analysis is one apartment, as they are measuring whether that apartment is vacant or not. The type of sample is stratified. The stratification allows the researchers to determine vacancy rate for different types of buildings. However, blocks within each planning area are chosen at random. There is no indication of whether the adding and subtracting of blocks is done at random as well. This means that the sampling is not necessarily a true random study. For the purposes of estimating the vacancy rate, this study method is acceptable, since vacancy itself is randomized. However, if only the blocks are randomized and the buildings are not, then this reduces the usefulness of the study in determining vacancy rates for specific buildings. This is because the choice of specific buildings is not adequately randomized. Without full randomization, even the overall vacancy rate is not entirely reliable. Thus, the sampling method can be improved by randomizing every aspect of the study.

c. The population in this study is welfare clients. The sampling frame is national. While the randomization draws from multiple states, the end figure in the study is not going to have this state breakdown. However, if average welfare assistance by state is derived, then the state would also be a sampling frame for this study. The unit of analysis is the individual, since at the end of the study a per individual dollar figure will be derived. The type of sample is random, because the individuals are chosen simply on the basis of their walking through the door, and the individuals did not know that the time they walked through the door would include them in the study. This study is adequate, if somewhat flawed. The thirteen states are not chosen at random, which leaves that process open to cherry-picking of states to be used for the data. There is still for bias in the state selection. The individuals, however, are randomized. The method is good because it is designed to be relatively free from bias and there is randomization of the individuals, but there is room for improvement because the process is not randomized throughout.