Sampling
Sampling is a process where a researcher selects part of the population to be used as a proportionate sample for a larger population. It is a very crucial method used for the collection of data and its core purpose is to enable a researcher draw the appropriate conclusions about a certain population from the sample collected. It involves the use of inferential statistics that gives us the ability to determine the characteristics of a population by way of observing the sample collected (Lameck, 2013).
Moreover, sampling is further divided into probability and nonprobability sampling. Probability sampling is a sampling approach where the subjects are given an equal opportunity to be selected; examples are simple random sampling, systematic sampling, and cluster sampling whereas nonprobability sampling is, a subject’s chance of being selected is not known; the examples are convenience sampling, consecutive sampling, quota sampling, judgmental sampling, snowball sampling etc (Lameck, 2013).
Delegate your assignment to our experts and they will do the rest.
Bias
Bias in sampling means that the data collected from a population do not bear the true distribution of the population due to some non-random reasons. While conducting human services research, it is important that one tries as much as possible to avoid bias of any kind (Lameck, 2013). Therefore, the representative sample should be provided by the respondents and they should be able to reliably and validly representing the opinions, thoughts or even feeling of the whole population. The outcome should thus be free of the researcher’s personal influence because it will render the sample invalid. The most used method of sampling to avoid bias is simple randomized sampling approach because other techniques can lead to biases findings depending on the population sample.
Data collection
Data collection is a technique used for gathering and also the measurement of information on the targeted variables in a systematic manner which will enable the evaluation of outcomes Luo, Y. (2009). There are four major scales of measurement in data collection that include:
Nominal scales: Which are essential in the labeling of variables without any form of quantitative value.
Ordinal scales: This categorize values in the order of significance but the difference between each of them is not exactly known.
Intervals scales: This are numeric scales where the order is known and the exact differences are also known.
Ratio scales Are the ultimate measure of values because they can tell us about the order and also the exact values between the values. Moreover, they have an absolute zero value which is essential for allowing a wide range of inferential and descriptive statistics.
Reliability
It basically refers to consistency in the data collected or the consistency in measuring tests. Therefore, the data should be reproducible. For purposes of making data reliable (Lameck, 2013), it is vital that a researcher generates failure data which is usually obtained from field studies or from panned reliability tests.
Some examples of tools for testing reliability are the Kuder-Richardson 20 which is a measure of internal reliability of binary tests i.e. right or wrong answers and the Cronbach’s alpha, which is a measure of internal reliability for tests which have multiple possible answers (Lameck, 2013).
Validity
Validity indicates just how sound a research is because it applies to both methods and design of the research validity in data collection. There are many tests for validity, for example, Concurrent validity, content, convergent, criterion, ecological, external, face, formative and predictive among others `
Importance of reliability and validity
Validity and reliability are important to ensure that results are not meaningless. Furthermore, they are useful (Lameck, 2013). in ensuring that the results are usable in measuring the results and they are able to answer the research questions. Nonetheless, the underlying reason is that these findings cannot be used to generalize all the findings; therefore, they are just a waste of effort and time (Luo, 2009). However, it should be noted that even if one study is regarded as valid in one case, it should not be generalized or assumed that it would be valid when measuring something else.
Data collection methods
Telephone surveys
Advantages
It facilitates for rapid data collection
The costs involved in this method are significantly low.
It can ensure that there is anonymity between the researcher and respondent.
It allows for large-scale accessibility of respondents
Disadvantages
There is less control over the kind of data collected.
There is lack of access to visual material to back up the data
Some of the needed phone numbers are sometimes inaccessible.
There could be problems with the answering machines, which limit the communication and collection of data.
The complexity of the research questions is limited.
Online surveys
Advantages
Drastically lower overhead costs
It allows for automation and real-time access of data
It allows for rapid deployment and return times
Respondents can participate on their own convenience
It also allows for design flexibility
Disadvantages
There could be possible cooperation problems among the respondents
Since there is no interviewer for probing, the data can easily be unreliable.
Since certain populations may lack access to the internet, there is a possibility of limited sampling and respondent availability.
Focus groups
Advantage
The main advantage is that the facilitator can easily gauge the respondent’s reaction to a particular set of data and in many cases can even suggest improvements on the end product.
Disadvantage
Compared to the other data collection techniques, focus groups are not as in-depth because they are unable to cover maximum depth regarding the topic of discussion because members may not entirely be honest on their deliberations.
Surveys via websites
Advantages
There is ease in data collection
Minimal costs in collecting data
It allows for automation of in data input and handling
There is increased rate in responses collected.
There is flexibility in design.
Disadvantages
The absence of an interviewer can lead to bias from the respondents
It is largely faced with the challenge of reach to specific populations like those that lack access to the internet
It allows for possible survey fraud because some predicate just to get participation incentive.
In summing up, my preferable data collection technique would be telephone surveys because they are an easier way of facilitating rapid data collection and the costs involved in this method are significantly low. Furthermore, the fact that it can ensure that there is anonymity between the researcher and respondent, the respondent can easily provide reliable and valid data and the disadvantages are not far-reaching (Lameck, 2013).
References
Lameck, W. (2013). Sampling Design, Validity, and reliability in General Social Survey. International Journal of Academic Research in Business and Social Sciences , 3(7).
Luo, Y. (2009). Using Internet Data Collection in Marketing Research. International Business Research, 2(1)
