UNIT #5 DISCUSSION
DQ #1: Data collection is one of the most challenging steps that prevent learners from completing their dissertation. What challenges do you foresee in collecting your data? What are some strategies you’ll deploy to overcome these challenges?

a) Challenges in collecting data
Reliability and validity
In data collection, reliability and validity of the data are important considerations, which a researcher should pay attention to. Reliability, in this case, can be described as the degree to which the data collection process is consistent. On the other hand, validity arises when variables measure what they were intended to measure. And, this, usually, happens when the data collection process has been reliable.
Data collection form
Construction of a data collection form is one of the important considerations that every researcher should consider, before the inception of the data collection process. Designing the form requires that a researcher pay a lot of attention to all the established variables, from the topic. If all variables in the research are not covered in the form, it might not be possible to get complete information to enable well informed conclusions from the data analysis.
The wording of the form is also very important. If the form is not carefully worded, it might bias the participants’ responses. The questions should also be impersonal to encourage open and honest responses. In the absence of this, it is likely that the full story would not be told. McNamara (2008)
The type of data to collect
At the planning stage for data collection, a researcher might be confronted with a problem of deciding the type of data to acquire. Collecting the right kind of data is an essential step in the process of ensuring that the research questions are answered, as well as in finding a solution to the problems identified. Failing to identify the right kind of data to collect might lead to an inconclusive study, or unintelligent analysis and conclusions.
Where will the data come from
After determining the type of data to collect, the next question should be: where will the data come from? The source of the data must be identified early, and the ideal time for doing this is at the data collection planning stage. Acting in this way, would ensure that time is used optimally, rather than wasted “chasing the goose”. If this part of the research is not properly planned for, chances are that questionnaires, interviews or surveys etc. might be directed to participants who do not know much about the topic under study. And, as such, the responses received, if ever they are received, may be of a questionable validity and, hence, not very useful in meeting the objectives of the research. Other than this, it is likely that the rate of response might be unacceptably low. Potential respondents might prefer not to respond on areas they don’t know much about.
However, in deciding where the data would come from it is necessary to keep an open mind so that sampling bias is not, unnecessarily, introduced in the study. So, sampling techniques such as random sampling, stratified random sampling should be carefully reviewed and used to deal with the potential problem of bias in the study.

Time
Data collection can take a lot of time as it, sometimes, involves going out in the field to meet the research participants and interviewing them. In the process, it is possible that the participants sampled for interview, as an example, may not be available at the agreed time and, hence, rescheduling might have to be arranged. In some cases, the data collection process might involve observation of the research phenomenon, which might have to be done over an extended period of time. This is particularly noticeable in longitudinal studies where comparisons and analysis of data collected over different time periods have to be made.
Cost
Research can be very expensive, and hence it is often necessary to ensure that funding is arranged in advance and, before the study begins. If financial resources are lacking, costs can become a limiting factor in collecting the right type and amount of data needed to conclude the research. Costs are, usually, incurred in travelling, accommodation and employment of data collectors, just as examples. b) Strategies for overcoming these challenges
Don’t discard original data
Maintaining the original data is important in case, among other things, the reliability and validity of the research findings are under question. Having such original data would become handy in case the researcher needs to go back and re examine the data, and to deal with doubts about the findings or conclusions of his or her report. Original data can also provide evidence that the results of the research were not manipulated; and if necessary verification can be instituted by a third party.
Furthermore, there is a common adage in research that there are two kinds of people. That is, those people who have already lost data and those who will lose data, Salkind (2005). So, given this truism, it is necessary that every researcher should ensure that a copy of the data file be kept, just in case there is an accident or mistake which leads to the loss of data. Having backup files is, therefore, a necessity hence backup facilities, such as: compact discs, memory sticks etc., should be acquired to prepare for such unfortunate eventualities.
Data collection form should be clear and easy to use
The data collection form should be clear and easy to use. The form should be expressed in a language that could easily be understood, by all potential respondents. Clarity in the data collection form can also help in ensuring that participants give informative and truthful responses. Further to this, the form should be easy to complete, since complexity might discourage potential respondents.
Don’t rely on other people for data collection
Collecting data can be a difficult job, as it, sometimes, involves travelling and the actual capturing of data. All these activities require hard work. There is, therefore, a temptation to delegate these tasks to somebody else, to assist. This often means employing people to do the work. And, it often happens that people doing this kind of work are remunerated on the basis of the number of data forms completed. As a result, the temptation to manipulate data in order to enhance earnings is common. When this happens, the validity of the research findings becomes questionable. On the contrary, there are times when people make genuine mistakes, but even in these situations the problem still need to be addressed. And, one of the ways to address such a problem would be to employ data verifiers to instil a sense of checks and balances.
Schedule where and when you will be collecting data
A schedule is a very important part of the research process. So, it is important that a researcher should schedule where and when data would be collected. Doing this helps in getting the research process organised and avoid wasting time. Clarity is achieved when each activity is time tabled with regard to when it will be performed. In addition, through scheduling a researcher ensures that important engagements are not inadvertently forgotten. Forgetting about appointments might lead to participants losing confidence in the researcher, and probably not willing to cooperate with him or her in the future.
Cultivate possible sources of participants’ pool
Who is going to partake in the research process is very significant, and the researcher should be prepared for any disappointment that might occur. The participants chosen for the research study are, usually, chosen using sampling techniques. And, it is likely that some of them might not be available at the time of the study. Contingency plans must, therefore, be put in place, just in case those participants originally counted on, are not willing or are not available to take part in the study. It is, as such, necessary to have alternative a pool of participants who could be drawn upon to participate in the study.
Follow up on participants who missed on their testing session/ interview
In case participants missed on their testing sessions or interviews, which is a possibility, then the researcher should be ready to make a follow up. This is a challenge that he should be prepared to face. And, as such, resources should be provided for this type of eventuality. Obviously, it is unlikely that participants will always be ready to partake in a research study at a time convenient a researcher. And, when it happens that participants are not available when needed, the first time around, follow ups should be made to ensure that their views are incorporated into the final report.

DQ #2: Why is it necessary to conduct descriptive statistics? Which measures are the most important from a statistical perspective?

a) Importance of descriptive statistics
Descriptive statistics is significant in terms of describing the data that has been collected. For data to be understood, one of the critical steps involves describing it and, this where descriptive statistics becomes imperative.
Descriptive statistics encompass measures of central tendency, such as the: mean, median and mode. These measures of central tendency are useful in understanding data; but, depending on the type of data, at hand, and the measurement method used. For example, the mean is often suitable for analysing ratios and intervals. The median, on the other hand, is suitable for the ordinal level of measurement; whereas, the mode is appropriate for the nominal level of measurement Salkind (2005).
Descriptive statistics also help in establishing the variability in the data. That is, measures of central tendency do not tell much about the type of data, on their own. Because of this limitation, it is usually necessary to complete the analysis by computing the extent of variations in the data. To this end, the range, standard deviation and variance are often calculated to aid in furthering the analysis and understanding of the data.

b) Important statistical measures
There are two important statistical measures, and these are the mean and the standard deviation. The mean is the most commonly used measure of central tendency. It is commonly called the average. It is easy to calculate, as it simply involves adding up all the values and dividing the sum by the number of values in a data set. The mean is, however, only considered a good measure of central tendency in the case of symmetrical distributions. For skewed distributions, it could significantly be affected by scores in the tail of the distribution. So, the mean represents a generalisation of values, and this means that its interpretation should be made with care.
Similarly, the standard deviation is the most widely used measure of variability or dispersion. It shows how much variation there is from the mean, hence the higher the standard deviation, the greater the degree of dispersion in the data. The standard deviation is also related to another measure of dispersion called the variance. In this case, the standard deviation is a square root of the variance. But, what also makes the standard deviation so outstanding is that it has multiple other uses; like in the measurement of confidence levels for statistical conclusions.

References
Salkind, N.J. (2005) Exploring Research. 6th edition.
McNamara, C (1997-2008) Overview of methods to collect information. Basic Guide to Program Evaluation. Minneapolis, MN: Free Management Library. Available at http://www.managementhelp.org/evaluatn/fnleval.htm#anchor1585345