What is descriptive research? How does it differ from the research types discussed in Chapter 9? Which is more suitable for administrative topics which generally involve specific problems within organizations? Explain.
Descriptive research is a type of quantitative research and is more appropriate for administrative research generally involving specific problems within organizations. This is due to the fact that descriptive research identifies characteristics of an observed phenomenon, and explore possible associations between two or more phenomena. It takes characteristics and converts them from perception to something quantitative and descriptive research examines a situation as it is. Research types discussed in Chapter 9 identify cause and effect relationships and considers many possible factors that might cause or influence a particular condition or phenomenon, with use of dependent or independent variables.
2. How do face-to-face and telephone interviews discussed in Chapter 8 differ from those discussed in Chapter 6?
According to Chapter 8, face-to-face interviews are structured or semistructured, enable the researcher to establish rapport with participants, yields the highest response rates in survey research, should not be limited in lengths, and the time and expense involved may be prohibitive. In Chapter 6 interviews are discussed during phenomenological research where these types of researchers depend almost exclusively on lengthy interviews with a carefully selected sample of participants. The phenomenological interview is often unstructured, differing from the structured or semistructured approach in Chapter 8, in which the researcher and participants to arrive at “the heart of the matter.”
3. In thinking about survey questions and associated rating scales, why are students discouraged from including all yes/no questions and encouraged to use checklists and rating scales that offer more responses, such as the Likert Scale?
It is useful to use checklists and rating scales to offer more responses. How surveys are structured is important because the kinds of scales on the questions for the surveys determine which statistics techniques to be used in analyzing the responses and data. Nominal, ordinal, interval, and ratio are all various types of scales which can be implemented for surveys.
4. For the Common Research Problem we worked on earlier this semester, describe the target population for a survey that would be used to understand the high capstone incomplete rate. If you are developing a sample, what considerations must you take into account, based on information presented in Chapter 8?
The target population all students who registered for the capstone class, MSA 699, for the past three years. To develop a sample, it will be important to take into consideration the nature of the data and determine which statistics, tools, and methods to use. Will it be a single group or multiple groups? Will a checklist be used in order to develop a list of behaviors, characteristics, or other entities under investigation? Or, will a rating scale be used when a behavior, attitude, or other phenomenon of interest needs to be evaluated on a continuum such as “never” to “always” which is an example of a Likert Scale. Which scales will be part of the survey, i.e. nominal, ordinal, interval, or ratio? We can use descriptive stats and inferential stats to enable us to look at our samples and infer from the sample how well it aligns with the entire population.
5. For your individual research project, assuming you were going to do a survey, which sampling method would be best suited for your study: simple, stratified, proportional, systematic or non-probability sample? Explain and justify your answer.
To research the use of career development resources and tools for my individual research project it would be best to implement a survey sampling method using the stratified sample which will require the researcher to take samples equally from each one of the layers in an overall population. Therefore, if there are ten different departments within Domino’s, stratified samples ensure that each department is represented in the survey data equally, yet still randomly and fairly.
6. How will you determine your sample size?
As a sample size increases, the distribution of sample means of size n, randomly selected, approaches a normal distribution. This sample size will be determined mathematically. In looking for a 50% response rate, and using a population of 150, I believe the sample size needed when entered into a formula states the sample size necessary is 80.
7. Having reviewed both qualitative and quantitative data collection methods, share with your fellow students which method you will use. Explain and be specific.
Quantitative will most closely represent the data I will collect for my research project and sampling is best suited for my particular study. I will create a survey for employees to complete in order to determine what the cause is for low career development resources usage, identify the survey objectives, identify my target population and sample size, the sampling methodology will be stratified in order to systematically, equally, and randomly select the sample. Quantitiative data collection methods will provide the best data to analyze in order to determine the answer to the questions.
Descriptive research is primarily concerned with what a situation is, not necessarily what is causing or the effects of the situation. It does not seek to change or modify the situation under investigation. Conversely, experimental research designs consider factors that might cause or influence a situation, condition or phenomenon. Dependent upon the specific research problem being researched, either descriptive research or experimental research may be more suitable. Descriptive research topics may examine correlations between variables in a situation while experimental research examines the variables that are causes and/or influences in sitations.
Qualitative research interviews such as those discussed in Chapter 6 tend to be informal, friendly, open-ended and revolve around central issues. In descriptive research interviews, the format is structured and emotionally neutral. Using yes/no type questions in an interview will not help to assess the degree to which people feel a certain way. For example, a strong objection to something cannot be captured and understood by a simple “no” response. Likert scale questions better capture the level of the yes/no answer which helps the researcher to recognize patterns within answers (e.g. strongly agree).
Identifying a target sample must be done in a random way that best represents a true sample of the population without bias. In the common research problem, it wouldn’t be a true representation of the population if only online students were sampled nor would sampling only those that completed the capstone project or only those that the researcher wanted to survey because they knew them. For my project “impact of merger on employee morale and productivity”, using the proportional stratified sampling makes sense. The layers of the organization would include individual contributors, managers and executives. However, since the individual contributors make up the strong majority of workers, a proportionate method would be the most representative of the population. This may include representative for every one executive, two managers and five employees. Based on the makeup of qualitative studies including the informal nature, open-endedness and looser structure, this is desirable for my method of choice. While I still will incorporate structure and a hope for emotional neutrality, I feel that it will be important to establish a rapport with the participants to increase participation and get to their true feelings on the issues at hand. One other thing to consider, however, is the human factor – that memory is not always accurate and to take that into account when analyzing participant answers.