Sampling is essential since it describe accurately specific parameters of a given population based on the description of the set of elements that are drawn from that population. Sampling fundamentals include the following principal factors: First is the population which is simply the collection of all the items that are considered to be of interest to the researcher. The sample is the representative part of the entire population. The sampling is considered a technique used in selecting an effective representative group from the larger population. Sampling error is a potential difference that occurs by chance between sample statistic values and that of the corresponding population parameter (Monette, Sullivan & DeJong, 2011).
There are two primary methods of sampling including probability and non-probability sampling. An example of probability sampling is random sampling where each item has an equal chance of being included. An example of a non-probability sample is quota sample. The main defining characteristics of this sampling are that the researcher sets the proportions of strata within the entire sample deliberately. According to Monette, Sullivan & DeJong (2011), when selecting samples for human services research, the researcher can avoid potential research bias by making sure that he or she uses specific methods that will enable him to eliminate possible sources of bias. Further, the researcher should measure the sources of bias and then include them as additional predictors within the research.
Delegate your assignment to our experts and they will do the rest.
Data collection is simply the process used by a researcher to gather and measure information on specific targeted variables in an established systematic fashion. This will further be used to answer key questions about the study, to test a hypothesis and evaluate the outcomes. In the data collection the researcher lays emphasis on the accuracy and honesty collection.
Scales of measurement are essential when conducting research and they include: nominal scale that deals with those variables that are believed to be non-numeric or those that have numbers with no value. The ordinal scale considers those variables where the order tends to matter however the differences do not matter. An interval scale is used to measure variables where the potential differences between the numbers matter a great deal. Lastly is the ratio scale which is considered to be the uppermost level of measurement. It tends to have characteristics of interval scale but with a static zero point (Monette, Sullivan & DeJong, 2011).
There are four types of reliability estimates. First is the inter-rater or Inter-observer reliability that is commonly used to evaluate the level at which various observers offer their regular estimates of a given phenomena. Test-retest reliability is utilized by a researcher to evaluate the level of consistency of “measure from one time to another” (Monette, Sullivan & DeJong, 2011). On the other hand, parallel-form reliability is utilized in the assessment of the level of coherence of outcomes of a two tests that are constructed in a similar way but from a similar content domain. Lastly, internal consistency reliability is used in the assessment of the level of uniformity of outcomes on items in a test.
There are several types of validity as applied to human services research: external validity analyzes the extent to which a given effect in research can be generalized to the entire population. Internal validity guarantees that the experiment design conforms to the established “principle of cause and effect carefully” (Monette, Sullivan & DeJong, 2011). Criterion validity evaluates if a particular type of test reveals a given group of abilities. Under criterion validity, we have concurrent validity that quantifies the test against a given standard test and the predictive validity that determines how sound a given test projects the abilities. According to Monette, Sullivan & DeJong, (2011), content validity measures how much a given measure will represent each element of a given construct. Construct validity is used to define how well a given test will measure up to its claim. Under construct validity, we have a convergent validity which measures that the constructs that are expected to be closely related are related and discriminant validity which is used to test that constructs that are expected to have no relationship have no connection.
It is important to make sure that data collection method and instruments are both reliable and valid in every research. This is because a reliable instrument used in research will generate consistent results irrespective of the study setting (Monette, Sullivan & DeJong, 2011). A valid instrument, on the other hand, will accurately measure what is to be measured by a researcher. A researcher will thus seek high internal validity to justify the statement and conclusion about the causal inferences in addition to the potential relationship the variables. An instrument that lacks validity will produce a skewed or even flawed result.
Telephone surveys entail a researcher conducting the survey on the population under study over the telephone. The advantages of this approach include: It ensures rapid data collection; it is cost effective and enhances the high level of anonymity. On the other hand, the disadvantages associated with this approach include: it allows less control, it has limited potential respondents, and the researcher might not access the telephone numbers.
Online surveys allow the researcher to collect data using surveys which are distributed to respondents online. The advantages associated with this approach include: It is believed to be highly cost effective. It consumes less time to gather data, it is also convenient to respondents and lastly it is believed to be highly flexible. The disadvantages associated with this approach include: it has a limited sampling in addition to respondent availability. Further, there are potential cooperation problems besides lacking the interviewer (Monette, Sullivan & DeJong, 2011).
Focus group is used particularly to gauge the public perception. The advantages of this approach area: enables the researcher to interrelate with participants hence can ask queries that will investigate more profoundly; the results are often easier to be understood. It provides information more quickly. However, disadvantages associated with this approach include: It is difficult to control group discussion hence need a lot of time. Further, the respondents might face peer pressure to provide similar answers.
Website survey is commonly used by businesses to understand and get public opinion concerning their brands. The benefits of this approach include: It can be used mainly to address sensitive questions since the interviewer is not present to ask questions directly. Further email is less costly and the fastest means to administer the survey to participants (Monette, Sullivan & DeJong, 2011). The demerits of website survey include: first issues has been raised that this approach can raise ethical concerns in addition to being intrusive. The email send can fail to be delivered or can be rejected as SPAM hence affecting survey.
In conclusion, telephone survey is the most suitable approach to be used when collecting information because it is considered one of the most efficient and rapid instruments in the collection of data. Further, it is highly flexible compared to the rest of the survey methods hence allows the researcher to change the questions and add more questions to gather in-depth information. Lastly, it is suitable because it is less costly besides enhancing a high level of anonymity hence allowing greater participation.
Reference
Monette, D.R., Sullivan, T.J., & DeJong, C.R. (2011 ). Applied social research: a tool for the human services (8th ed.). Retrieved from The University of Phoenix eBook Collection database
