Describe how you would find a sample that would represent your population of int

Describe how you would find a sample that would represent your population of interest?
Population of interest is based on Population from PICO (T): DIabetes Mellitus type 2
It is important to obtain a sample that is representative of the target population. A sampling plan enables the researcher to set inclusion and exclusion criteria in an attempt to limit error and bias. Since quantitative researchers are not interested in rich verbal description, but instead desire to generalize findings to a greater population, two main sampling designs are used. The first type of method is the probability, or random sampling, method.
This type of sampling method attempts to reduce sampling error through randomization or random selection. This means that every subject or participant has a probability higher than zero of being selected for the sample. This is not to be confused with random assignment, which has nothing to do with the selection of subjects. Random assignment is the process of randomly allocating subjects to different treatments in experimental designs. We will review the following probability sampling methods: simple random, stratified random, cluster, and systematic.
Simple Random Sampling
This method of sampling is very basic, but entails a great deal of work. First, the sampling frame must be established. This is the actual list of elements from the population. After the list is established, the researcher needs to randomly select the desired number of elements for the sample. This can be accomplished by using a computer, a random numbers table, or any technique the researcher desires to use. Examples include: writing names on slips of paper, placing them in a container, mixing them, then drawing out one name at a time until the desired size is reached. If equal opportunity of being selected is desired, then selection with replacement is used. A random numbers table can also be used. The researcher closes his eyes and places a finger or a pencil anywhere on the table. This is the number where he will begin his selection. He can then move up, down, right, or left to choose the sample. Unfortunately, it is difficult to obtain all the elements of a sampling frame, therefore other sampling methods are used.
Stratified Random Sampling
In this form of sampling, the researcher desires to achieve an even distribution throughout the sample, so the sample is divided into strata or groups. The variables chosen must be correlated with the dependent variables. Variables commonly stratified are age, gender, ethnic background, socioeconomic status, different institutions, types of care, and diagnosis. The stratum can be proportionately or disproportionately selected. For example, one may want an equal number of white, black, and Hispanic participants represented in the sample. Stratifying can decrease sampling error, increase power, and increase representation.
Cluster Sampling
There are times when simple or stratified sampling cannot be accomplished due to travel time or cost. Large-scale studies and national surveys need to use cluster sampling: ” a sampling frame is developed that includes a list of all states, cities, institutions, or organizations (clusters) that could be used in a study, and a randomized sample is drawn from that list” (Burns & Grove 691).
Systematic Sampling
This form of sampling involves the selection of every kth case from some list or group, such as every 10th person on a patient list. First the population size is divided by the desired sample size, which provides the interval width. The interval is the distance between each element. For example, if the population size is 1000 and the desired sample size is 100, then 1000/100 = 10. Thus k =10 so every 10th element in the sampling frame would be included in the sample. A number table selects the first element randomly, then every 10th element would be included.
3. Non-Probability Methods
Non-probability sampling does not provide an equal opportunity for each element to be selected for the sample. These sampling methods increase the chance that they will not be representative of the population. However, most nursing studies use non-probability methods. Convenience and quota sampling are used more in quantitative designs, where purposive, network, and theoretical methods are used in qualitative designs.
Convenience and Snowball Sampling
In this method, the researcher uses the most available participants. People who fit the inclusion criteria and are readily available are recruited for the study. For example, students in the university where the researcher teaches, patients in a local hospital, or participants in a support group are often utilized for this kind of study. The researcher has access to obtaining the participants. Snowballing occurs when members recruit other members based on their knowledge of inclusion criteria. Convenience samples are used for descriptive and correlational studies. They are inexpensive, and require less time for acquisition. Also, specific topics that may be difficult to study can be accomplished by using a convenience sample. However, the risk of bias is great and cannot be evaluated.
Quota Sampling
Strata are first identified, and then the researcher determines how many participants are needed for each stratum. A convenience sample could under- or over-represent a specific population. A quota sample can assure an appropriate number of elements for equal representation. This is very similar to stratified random sampling. However, the quota method also shares the same biases as convenience sampling, but is also an improvement by providing improved representation.
Purposive Sampling
This form of sampling is also called judgmental or selective sampling. The researcher is knowledgeable about the characteristics and sample criteria necessary to study the phenomenon. Therefore, the researcher handpicks participants to be included in the study. Qualitative researchers use this method so the richest descriptive data can be obtained.
Theoretical Sampling
This method is used in grounded theory research. Any person or group that can provide information promoting or contributing to theory development is sampled. The purpose is providing relevant data for theory generation.
4. Sample Size
A power analysis is conducted to determine a sample size for a quantitative study. Power is the capacity of the study to detect differences or relationships that actually exist in the population (Burns & Grove, 2013, p. 714). It is the “probability that a statistical test will detect a significant difference or relationship that exists, which is the capacity to correctly reject a null hypothesis” (Burns & Grove, 2013, p. 714). The minimum standard of power for a study is .80. A power analysis determines the sample size needed to achieve power. In order to determine the sample size and maintain power, the significance level, or p value, must be set, and effect size must be determined. Kraemer & Thiemann stated that a larger sample size is necessary when a more stringent significance level (.001) is established. Thus a p value < .05 is the best choice. Effect size involves the existence of the phenomenon of interest in the population. Cohen classified effect size as being small .2 to .3, medium .4 to .6, and large .7 to .8. Large differences are easier to detect than small differences. Therefore, if the effect size is .8, then a smaller sample can detect the difference. A small effect size would need a large sample in order to detect a difference. For example, anxiety the night before surgery would be higher than one week before surgery. Therefore, it would take a small sample to detect anxiety the night before and a larger sample to detect a difference the week before. Evidence from previous research studies assists the researcher in determining the effect size. Some basic factors useful in determining sample size established by Kraemer & Thiemann (cited in Burns & Grove, 2013, p. 357) are as follows: The more stringent the significance level, the greater the necessary sample size. Two-tailed statistical tests require larger sample sizes than one-tailed tests. The smaller the effect size, the larger the necessary sample. The smaller the sample size, the smaller the power of the study. It is unfortunate that many of the studies conducted in nursing do not use power analysis to determine sample size. More nursing researchers need to use power analysis to determine the sample size necessary to detect significance within the studies. 5. Recruitment and Retention Once the sample size has been determined, the sample must be recruited and retained so power is not compromised. Participants must feel valued and respected. The researcher must present a positive approach, and be pleasant and informative. The participant should not feel pressured to be in the study. Body language and words need to be non-threatening. If different ethnic groups are included, the researcher should be culturally competent in communicating with the participants. Other data collectors besides the researcher must be educated in following the sample plan, so as not to compromise random sampling or interject bias. When questionnaires are mailed, creativity is required to strategize on the presentation of the information packet, token of appreciation, and follow-up. Letters and cards mailed as a follow-up assist in raising response rates. Timing, however, is important. The researcher does not want to wait too long before sending the card in fear that the questionnaire might be thrown away, and too short of a time frame and the participant might get annoyed. Recruitment of participants can be accomplished through use of media, newspapers, and agency newsletters. Radio and public service announcements, church bulletins, and posters placed in neighborhood stores can be used to recruit participants. Clinical trials require different strategies. Recruitment may be occurring at several sites, participants may be screened twice, and screening logs are kept to record data. Contact must be maintained with the participants on a regular basis. All participants must be well informed and encouraged to continue their participation in the study. As always, non-aggressive recruitment strategies should be used. Retaining participants is equally important. Mortality can affect the power of the study if sample size decreases from what is required. Sometimes people move away, so obtaining and updating addresses and telephone numbers, including those that are unlisted, is important. If studies are longitudinal, then bonus payments may be included for completing a phase of the study. Care must be taken to not compromise the voluntary nature of the study, or exploit the participants in any way. Remember that the participant's time is valuable, and should not be taken for granted. Make sure data collection is conducted in a nurturing, pleasant environment. A researcher who is perceived as being altruistic, ethical, and nurturing has a better chance of retaining participants.Introduction This module discusses the important aspects of sampling. Population and sample are defined, and differences between the two are examined. Probability and non-probability methods are explained, along with time frames, sample size, sampling error, and sampling bias. Guidelines for critiquing the sampling plan are also reviewed. All other APA formatting guidelines should be followed. For example, in-text citations must be formatted with the appropriate information and in the correct sequence (Author, year), reference list entries must include all appropriate information following guidelines for capitalization, italics, and be in the correct sequence. Refer to the APA Publication Manual 7th ed. for each source type's specific requirements.