Introduction
In this research, 3600 diabetic patients were surveyed from twelve hospitals, but due to exclusion criterion, only 1,200 were considered for this particular research. Out of the tools provided, I applied randomizer in selection of the participants. This tool categorized the whole population in to 12 sets with 300 participants each. The system allocated numbers to the participants out of which 100 were picked randomly.
Research question
Do the diabetic patients benefit from sharing knowledge on diabetic condition with the physicians?
Inclusion criteria
- Only patients whose age ranged from 40 to 65 were included. This age category represents the healthy class not susceptible to any ailment.
- Patient diagnosed as having either diabetes type 1& 2 for 2 years prior to research. A period of two years gives a clear confirmation for those suffering from diabetes.
- Those having HbA1c above 8% (HbA1c refers to Glycosylated hemoglobin, a measure of long terms sugar levels). Normal sugar levels range between 1-8% and anything above 8% gives a clear indication of diabetes case.
Exclusion Criteria
- Patients, who prematurely interrupted, discontinued or had dose adjustment due to tolerability issues are excluded because they may have sugar levels which may give distorting information.
- Patients who have been treated for Hepatitis C Virus within 3 months prior to screening. Hepatitis C therapy is known to affect the body sugar levels and in most cases causing them to rise beyond the normal levels.
- Known HIV infection or chronic Hepatitis B Virus infection. These chronic problems are known to cause irregular changes in body sugar levels as they are known to drain the stored glucose and. The patient may in turn take much more glucose causing irregular sugar levels.
Reasons why the sample is a representative of entire population
Sample representativeness basically expresses the degree to which a given sample data precisely and accurately represents a characteristic of the entire population (Davis, Wagner & Groves, 2000). Therefore, through application of scientific methods, the entire sample proved to be a true representative of the entire population as the selection procedure was free from bias. The sample size was also sufficient and efficient in relation to the whole population, and more so, the method used in selection in all the twelve hospitals was consistent.
Generalizing the sample results to the entire population
Researchers cannot make observations for every individual in the population under study and as such they collect data from a sub set of individuals and use such observations to make inferences about the entire population (Denzin, 1994). Ideally, the sample represents the entire population if it bears the characteristics under interest. For that matter, the researchers’ conclusion becomes applicable to the entire population. Many strategies can be used to create a probability sample which defines the target population from which the sample is drawn and to which the sample data will eventually be generalized. Systematic sampling method was used to divide the entire area under consideration after which random sampling was used to recruit the participants. Thus the sample is a true representative of the entire population and as such, the sample results can be generalized to the entire population.
Determining if the sample size is large enough to represent the entire population
Besides using the appropriate methods and considering participants adequate response rate, the sample size must also be evaluated to determine the representativeness of the sample under study. Important to note is that smaller samples (those with fewer than 1,000 respondents) bears a greater sampling error unlike larger samples (Cooper et al., 2001). Sampling error refers to the number that describes preciseness of an estimate from any one given sample. Therefore, considering that the sample put under study has 1,200 respondents, it can be taken to be large enough to form a good representative of the entire population.
Expression of sampling error:
e = z√ (p %( 100-p %))
√ s
Where:
- e = sampling error (proportion of error that is acceptable)
- s = sample size
- z = degree of confidence
- p = estimate of the proportion of patients falling into the group suffering from diabetes
Applying the sampling error formula at 95% confidence level, sample of 1,200 respondents with an estimate of 20% suffering from diabetes is calculated as follows.
e = 1.96√ (20(80))
√ 1200
= +/- 2.6%
This means that, based on a sample of 1,200, we can be 95% sure that the true measure from the whole population from which the sample was drawn will be within +/-2.6% of 20%, that is, between 17.4% and 22.6%. This range is not very large and, hence, the sample size is large enough.
Ways of determining the sample size
It is always important to determine the sample size failure to which a researcher may end up with a sample that is too large, which may waste resources and time (Cooper, 1994). However, the sample may also be too small such that it leads to inaccurate results. To determine the minimum sample size needed to estimate whether the selected sample is a true representative of the entire population, a researcher must determine the below listed parameters:
- Population mean
- Sample mean
Notably, the above two parameters should yield a similar figure but in most cases they differ due to the margin of error which the researcher ought to have established the acceptable level before carrying out the research.
References
Cooper, H. (1994). Laying the foundations of diabetes care. Practice Nurse Journal, 7(5), pp.78–82.
Cooper, H., Booth, K., Fear, S., & Gill, G. (2001). Lessons from chronic disease patient education. Patient Education and Counselling, 44(15), pp.107–117.
Davis, R. M., Wagner, E. G., & Groves, T. (2000). Advances in managing chronic disease. British Medical Journal, 320 (27), pp.525–526.
Denzin, N. K. (1994). The art and politics of interpretation. Handbook of Qualitative Research. 49(7), pp. 500–515.