Descriptive Research Design And Data Collection: A Case Study

Introduction to Descriptive Research Design

Discuss about the Logistic Regression Limitations of Model.

Descriptive research design usually describes the features of the people, object, groups, environment and the organizations. Descriptive research mainly tries to describe the given situation. This type of research is conducted after the researcher has gained all the necessary information of the situation that has to be studied. The information will be collected using the exploratory research and the findings will be developed and made understandable with the help of descriptive research design.

• The main conditions of the facts in a case that is under the study can be ascertained by the descriptive research design.
• This research mainly provides a qualitative or quantitative or both qualitative and quantitative measures of the case that is under the study.
• The cause of the main conditions is not under the importance.
• The conditions of the case may be studied at different time-points and the changes that has taken place in between these time-points should be noted for the purpose of evaluation.
• The characteristics between two different groups must be evaluated to identify their differences and their similarities.
• Control of the variables or its conditions is not done usually in descriptive research designs.
• Besides case studies, all the descriptive studies are cross-sectional. Thus, descriptive research design usually studies the different sections involved in the same group.
• The studies that are usually conducted on the main conditions of the case that is under the study can be repeated in order to verify or compare the results with the previous results.

The current case study is mainly focused on the styles in which decision is made for the students of the university while purchasing sports shoes and mobile phones. The decision is made on the basis of the following three reasons.

• The students of the university are young adults and they have buying powers. They also have the potential so that they can increase their buying power.
• The most important market of the mobile phones is given by the students of the university.
• Various styles of decision-makingsare related to different types of retail businesses.

Since mobile phones and sports shoes are common between students as well as serve as a major part of fashion and technology, it is important to make the right decision for these products. The main objective of this research is to identify whether the stronger decision making style of the foreign brand increases the possibility of staying with the foreign products. To do the research, data has to be collected and then analyzed in order to understand on what basis the students are making their decisions. Thus, descriptive research design is the most appropriate research design for this study.

Are the methods used to collect data (e.g. questionnaires – Appendix 1) sufficiently justified and described? What further information (if any) do you require?

The most important method of data collection is by preparing a questionnaire. A questionnaire contains a lot of questions related to the research topic. These questionnaires are sent to the selected respondents are asked to fill out the forms.

A questionnaire must follow the following principles at the time of its construction.

• The items included in the questionnaires must be relevant to the research objectives.
• The participants required for the research must be understood properly.
• Easy and simple languages must be used for the questions in the questionnaires so that the questions can be easily understood by the participants.
• The items that will be mentioned in the questionnaires must be simple and easy for the participants to understand.
• “leading” questions and “loaded” questions must be avoided.
• Double-barreled questions must be avoided.
• Double negatives must be avoided.
• It is important to determine what type of question is necessary – open ended questions or closed ended questions. Most of the questions must be closed ended. Otherwise there will be a problem while analyzing the results.
• The categories of responses that are required for the closed-ended questions must be considered.
• Measurement of abstract constructs must be computed by using multiple items.
• The questionnaire must be prepared in such a way that the participants must find it easy to use.
• Pilot survey must always be conducted before using the questionnaire for the research in order to test the reliability of the questionnaire. If the questionnaire is reliable, then only it should be used to collect the data for the research. If it is found unreliable, then the questionnaire must be developed.

Keeping the above principles in mind, the questionnaire developed in the research that has to be analyzed. It can be seen that the questionnaire developed for this research contains all closed ended questions. The language of the questionnaire is simple and all the questions are quite relevant to the topic. The questionnaire contains two parts, the first part is the demographic part and the second part is the hypothesis development part. There are no double-barreled and double negative questions in the questionnaire. Thus, it can be said that the questionnaire used in this research is quite justified and described properly.

The Case Study: Decision-Making Styles of University Students

All the information that should be important in conducting this research on using mobile phones and are provided in the questionnaire. Thus, no further information is necessary. Among all the types of data collection processes, data collection with the help of questionnaire is the most effective.

In this research, 628 surveys were distributed electronically. The main population of interest or the target population was the undergraduate students in Guangdong. Among the 628 questionnaires, 584 questionnaires were found valid. Thus, analysis was conducted on the basis of the 584 valid responses. The success rate of the questionnaire in this case is 92.99 percent. The data that was selected in this research was by using the method of simple random sampling. The method used can be said to be simple random sampling as the population of interest, that is, the undergraduate students of Guangdong, China has an equal probability of being selected in the sample. The questionnaire was distributed electronically and all the students had access to the questionnaire. From them some responded and others did not. Everyone had the same chance of answering the questionnaire.

Sampling is broadly classified as subjective and objective. In subjective sampling, the selection of members depends on the personal judgment or discretion of the sampler, while in objective sampling, there is a specific rule of selection and the selection is independent of the judgment or discretion of the sampler. Objective sampling, in its turn, is classified as non-probabilistic, probabilistic (or random) and mixed sampling. When, there is no probability assigned to the mode of selection but the sampling is done according to a specific rule, then the sampling is called non-probabilistic. In random sampling each member has a definite pre-assigned probability of being selected. Thus, if a sample is drawn from a population in such a way that each member of the population has a definite pre-assigned probability of being selected in the sample, then the sampling is called random sampling or probability sampling. In this research, each undergraduate student had an equal probability of being selected in the sample. Thus, the sampling method used in this research is random sampling.

There are certain strengths of this type of sampling methods. The first one is that a very little amount of knowledge about the group of study population is required before doing the sampling. Secondly, in this type of sampling, classification errors are not present. Thirdly, the use of inferential statistics is appropriate in this type of sampling. Fourthly, this sampling method is completely unbiased and is the simplest method of sampling. Fifthly, the sampling errors can be assessed very easily in this method. Sixthly, the sample obtained with the help of this method represents the whole population. Thus, it can be said that, the sample that has been selected for this study is free from classification errors, represents the whole population and is completely unbiased. The sample of 584 undergraduate students is supposed to represent all the undergraduate students in Guangdong and thus the measures obtained from this sample will be an estimate of the total population of undergraduate students.

Data Collection and Justification

There are certain weaknesses of this method. A desired sample size is necessary for this type of sampling in order to make it a proper representative of the whole population. If the population units are heterogeneous, random sampling cannot be used. In this case the population units are homogeneous, thus it is not a problem to use random sampling in this research.

In this research, at first exploratory analysis of the factors have been done. With the help of this analysis, the variables which are the most relevant for the study and the variables which are less relevant to test the hypothesis has been identified. The association of a variable with all the other considered variables has been computed in this part. After this exploratory analysis, logistic regression has been run to test the hypothesis that has been framed in the literature review section. Whether the decision making styles influence the intention of switching of brands for the customers has been tested using the logistic regression.

The statistical method which is used to analyze a dataset where the dependent variable can be estimated or determined with the help of one or even more than one explanatory or independent variable. Usually in a logistic regression, the outcome or the dependent variable is discrete or binary variable. The variable can take only two values – 0 and 1, with zero indicating the absence of the event and 1 indicating the presence of the event. In this research, the outcome variable is switching of brands. If a student has an intention of switching brands, it is coded as 1, and a student with no brand switching intention is coded as 0.

The main intention of logistic regression is to determine the model which will be the best fit for the binary character that has to be predicted on the basis of the independent or the explanatory variables. The degree of influence of the independent variables is determined from the logistic regression. Which factors significantly influence the outcome variable and which factors are insignificant are also determined from this regression model. Thus, to sum up, a logistic regression does not minimize the error sum of squares in a model. Instead, it maximizes the likelihood of occurrence of the sample values by identifying the parameters that are responsible for the maximization. In this research, the variable gender is significant in the brand switching context of mobile phones. The male people prefer local brand mobile phones while the females prefer foreign brands. The condition is just the opposite in terms of sports shoes. Quality consciousness has also been significant for the purpose of brand switching. There is always a brand preference for most of the customers. Consciousness about fashion and recreation also has an influence in the switching brand context. The price of the products and also the multiple options that are available for the customers have proven effective in influencing the brand switching intention. These have been determined from the logistic regression.

There are certain limitations of the model. The outcome variable has to be binary in order to run this model. Presence of outliers in the model does not give proper predictions. The presence of multicollinearity between the predictor variables affects the results. This is the reason for running the correlation test before the logistic regression.