Inferential Because it Makes Claims

¶ … inferential because it makes claims about the population of adult Americans based on a sample of 9000 persons. The use of a sample proportion to estimate the population proportion makes this study inferential (Gravetter & Wallnau, 2008).

The research question in the study was; does the lack of health care increase the risk of death?

The data were obtained using a survey of 9000 persons tracked by the U.S. Centers for Disease Control and Prevention. While it is not explicitly stated in the article that the data was collected using a survey the tracking of individuals could not be done using an experimental design. Additionally, the purpose of the study suggested that it would engage in a correlational approach to explicating the problem.

The exclusion of persons aged 65 and over is an attempt to eliminate bias, as the inclusion of these persons would create a systematized form of error within the study. These older Americans receive health care through Medicare.

5. The conclusions drawn from the article are warranted because they are logical. Firstly persons who do not access medical attention will die from aliments that require medical attention. Secondly, the design of the study and the sample used are representative of the country and it would be legitimate to use such a sample. Finally, the design of the study followed a similar study done in 1993, therefore there is methodological support for the approach employed.

6. A large trial is necessary to ensure that the sample would be representative of the population. This representativeness means that the sample is similar to the population in key characteristics and the is little difference between the sample proportions and the population proportions. The error in the sample is therefore small.

7. A control group is needed to ensure that non-spuriousness is address adequately in the study. Using a control group means that the researcher is certain that the independent variable has the stated effect on the dependent (Lenth, 2001).

8. The double blind feature takes care of the propensity of human error to seep into the study. When both the participants and the researcher is unaware of which group is the treatment group, other variables that can have an effect on the study are controlled for.

9. The use of volunteers would have biased the results as that would have introduced systematic error into the study. The generalizability and validity of the study would be called into question (Creswell 1994). It is only through randomization that random error can be statistically determined.

Chapter 2

10-B

11-B

12-B

13-F

14-F

15-T

16-T

17

A random sample is similar to a convenience sample and a systematic sample only as they select member of a population for investigation. All methods of sampling will contain error. It is different because with a random sample there is a known probability of selection while with convenience sampling it is unknown.

18. Stratified sampling could be used where the population contains characteristics that can be placed together to construct a strata. In a clinical trial cluster sampling could be used to select individuals from a similar geographic area.

19-a

20-B

21-B

22-C

23-B

24-C

25-B

26-B

27-B

28-C

29-a

30-a

31-a

32-a

33-a

34-a

35-B

Quiz 2

1-True

2-False

3-True

4-False

5-False

6-True

7-True

8-False

9-True

10-False

11-True

12-True

13-False

14-False

15-False

16-B

17-D

18-C

19-a

20-D

21-C

22-B

23

H0 is the null hypothesis and it generally states that there is no difference or no change. H1 is the alternative hypothesis or the researcher hypothesis and is generally the opposite of the null hypothesis. Hypothesis testing centers on the acceptance and rejection of the null hypothesis.

24.

In most clinical research studies the researcher would use a one-tailed test. This is so for two basic reasons the first is that one tailed tests are more powerful and the second is that theoretically the researcher is usually aware of the direction in which the change is expected to take place.

25

The alpha level for this treatment would be .01. This level of significance is appropriate because you would want to ensure that the treatment is actually working. A lower level of significance would mean a higher risk of rejecting the null when it is true (Aron, Coups & Aron 2011). This would mean a greater risk of putting a product on the market that does not work. .05 is too high a risk to take, for replacing the standard treatment.