Data Collection And Data Analysis Physical Education Essay

The research method plays an essential role in addressing the research objectives with reliable and valid data. This section illustrates how and why a certain approach chosen to answer the research questions.

The methodology of this research bases on “research onion” model (Saunders, Lewis and Thornhill, 2012). The structure of this chapter follows the layers of the research onion model.

Figure 0.: The research ‘onion’ – Source: © Mark Saunders, Philip Lewis and Adrian Thornhill 2008 (Saunders, Lewis and Thornhill, 2012, p.108)

3.1. Research philosophy

Selecting the research philosophy is a necessary stage in the research process because it is important to reflect the perspective of a researcher and it also influences directly on the choice of research strategy and research methods (Saunders, Lewis and Thornhill, 2012). In business and management research, the popular philosophies includes positivism, realism interpretivism and pragmatism (Creswell, 2008; Saunders, Lewis and Thornhill, 2012). Thus, this research should determine the suitable philosophies based on the philosophy theory and the research questions and research objectives stated above.

Get Help With Your Essay

If you need assistance with writing your essay, our professional essay writing service is here to help!
Find out more about our Essay Writing Service

Regarding the theory philosophy, Saunders, Lewis and Thornhill (2012) discuss that there are two key points of view including ontology and epistemology. Firstly, based on ontology, the nature of reality or being of researcher’s view about the positivist is objective, external and independent with social factors, but researcher’s view about the interpretivist is subjective, social constructed and changeable depend on the participants (ibid.). On the other hand, on the way of thinking about epistemology, the valid knowledge of researcher’s view about the positivist focuses on causality, ‘reducing phenomena to simplest elements’, data and facts; but researcher’s view about interpretivist concentrates on social phenomena, feelings attitudes and detail of situations and subjective meaning (ibid., p.119). These viewpoints between ontology and epistemology are difference, and each of them will impact on the way of thinking about the research process. Additionally, the pragmatist research philosophy tends to be the combination of both positivist and interpretivist.

The positivist philosophy is appropriate choice for this research owing to the research objectives to examine the relationships among measurements of an e-learning system success and the using e-learning system of students to support their KM. Regarding ontology view, the e-learning system is an objective and singular. Furthermore, collecting data to analyse the e-learning system success in this research proves that it is positivism based on epistemology view. However, investigating students’ attitude can use interpretivism but it is reasonable with positivism.

3.2. Research approach

Due to positivistic philosophy and the research questions to test theory and the prior researches discussed in the literature review chapter, this research is suitable with deductive approach. This approach includes 5 progressive stages: (1) inferring hypotheses from the theory (test relationship among variables in the e-learning system success); (2) proposing the relationships among variables in this research (e.g. users’ attitude positive impacting on system acceptance); (3) “testing operational hypotheses” (using statistical software to estimate hypotheses); (4) discussing the outcome to confirm the theory; and (5) altering the theory based on the findings (Robson, 2002 cited in Saunders, Lewis and Thornhill, 2012, p.124-125).

Furthermore, the detailed progress of this research based on this deductive approach is illustrated in figure 3.2 (Maylor and Blackmon, 2005).

Figure 0.: The deductive approach process (Maylor and Blackmon, 2005,p.56)

3.3. Research strategy

Creswell (2008) stated that research strategy significantly influence on the direction of the research, as a result, the choice of research strategy is essential in research progress. The factors impact on the selected strategy including the research questions and objective, the choice of research philosophy, research approach and other resources such as time constraint, finance or data access (Saunders, Lewis and Thornhill, 2012).

Due to the selected deductive approach, experiment, survey, ground theory or case study are the research strategies that can be applied for this research. Experiment strategy tends to concentrate on a specific group. Moreover, case study is often used in specific research for a period of time. Hence, both experiment strategy and case study are not suitable for this study due to of the research questions. Because this study is developed on the previous researches in difference context, the ground theory strategy is not appropriate with this research. On the other hand, the survey approach is the most rational choice for this research within limited time. Survey strategy can be defined as a specific sampling from the population and the structured questionnaires are designed to test theory (Malhotra and Birks, 2007). This strategy is suitable to test the relationships between variables in research objectives using quantitative data method (Saunders, Lewis and Thornhill, 2012). In addition, this survey strategy also appropriate with cross-sectional time horizon studies (Easterby-Smith et al. 2008; Robson 2002 cited in Saunders, Lewis and Thornhill, 2012). Alternatively, the survey strategy also has some limitations such as irrelevant or inaccurate responses of the questionnaire or possibly insufficient sample. The number of responses may not achieve the target of at least 95% of confidential level because people are not able or willing to answers the questionnaire (Girden and Kabacoff, 2010). Understanding the potential drawbacks of the survey strategy is vital that result in well preparing in data collection plan.

3.4. Research choices

Due to the selected positivistic research philosophy as well as deductive approach, the mono method with quantitative approach is the research choice of this study. Using mono method seems to be adequate because this is not ground theory and experiment research strategy. Additionally, referring the section 2.2 of the literature review chapter, the mono method with only quantitative research approach tends to be used in almost previous researches to measure the e-learning system success and to test the research model. Hence, the theory and research framework of based on previous researches are tested in this research in the context of the University of Southampton with the quantitative approach.

3.5. Time horizon

This research has been conducted in three months. Thus, it is appropriate with cross-sectional time horizon owing to time constraint. Cross-sectional researches are appropriate to study specific phenomenon at specific time while longitudinal researches are suitable to study change and development over a long period of time (Saunders, Lewis and Thornhill, 2012).

3.6. Research design: Data collection and data analysis

This section aims to describe detail about the quantitative approach to collect and analyse primary data with sampling method, questionnaire design, pilot testing, data collection and data analysis sub-sections.

3.6.1. Sampling method

The context of this research is the University of Southampton. Therefore, all students and alumni of the University who has used the e-learning system at the University can participate on this research. Due to applying survey research strategy in this research, the most suitable sampling method for this study can be probability samples. Based on the probability sampling method (another name is representative sampling), the research questions and objectives can be achieved by evaluating “statistically the characteristics of the population from the sample” (Saunders, Lewis and Thornhill, 2012, p.213). In this research, the population which is all students using the e-learning system at the University of Southampton is generalised from the sample which is the students participating in the questionnaire.

The most appropriate sampling technique for this study is simple random sampling but the sampling frame size and the cost and time consuming of this sampling technique is high (Saunders, Lewis and Thornhill, 2012). Hence, multi-stage can be used as the sampling technique to save time and cost. In the first stage of this sampling technique, the list of schools at the University of Southampton is drawn. Then, in the second stage, a simple random sample of students can be selected by chosen schools. In fact, the selected schools may be not random because it is not acceptance to help of all schools at the University to broadcast the survey to their students. This study uses online questionnaire as a result of sharing questionnaire easily via emails or social network (such as Facebook). Indeed, several schools at the University are willing to help sending the questionnaire to all their students’ email.

Higher Education Statistic Agency summary that there are more than 23,000 students at the University of Southampton in 2010/2011 (HESA, 2012). Thus, the population in this study can be more than 100 thousands because both students and alumni has been used the e-learning system at the University since at least 4 years. This research targets to achieve 95 per cent confidence level and 5 per cent margin of error. Thus, referring to figure about ‘sample sizes for different sizes of population at a 95 confidence level’, the minimum sample size is 383 (Saunders, Lewis and Thornhill, 2012, p.219).

3.6.2. Questionnaire design

Designing a questionnaire is a vital stage in data collection technique to support positivism methodology, deductive approach and survey strategy in business and management research (Saunders, Lewis and Thornhill, 2012). The questionnaire technique is used to test the reliability and validity of hypotheses proposed from research objectives and research framework (Neuman and Neuman, 2011). There are many benefits of using questionnaire technique in research. For instance, it is an effective and economical approach to collect primary data (Patten, 2001). It also provide clear result in tabular to analyse easily (ibid.). Moreover, it can be managed anonymously and asynchronously (ibid.).

The type of questionnaire in this research is self-administrated online questionnaire due to the convenience and effectiveness of broadcasting as well as preparing data. The online questionnaire is designed on the iSurvey platform which is endorsed by the University of Southampton (www.isurvey.soton.ac.uk). Because iSurvey is a high quality tool and many advantages such as no cost, secure, friendly interface, ease of use, reliability and stability, it is used to design questionnaire by almost students of the University. Following the ‘questionnaire research a practical guide’ of Patten (2011), the designed questions are clear, short, simple and avoided common errors. Due to the willing help and limited time of participants, only necessary questions are presented in the questionnaire. Moreover, the questionnaire is designed with a short time to complete (around 10 minutes).

The validity of questions in questionnaire can lead to accurate data after collecting, and the reliability means the consistence of collected data (Saunders, Lewis and Thornhill, 2012). Bourque and Clark (1994, cited in Saunders, Lewis and Thornhill, 2012) stated three effective methods to design individual questions that are adopting questions used in other questionnaires; adapting questions used in other questionnaires; and developing own questions. Thus, rich literature review significantly supports questionnaire design with high reliability and validity because the questions in questionnaires are tested in the previous researches. Moreover, clear questions in questionnaire are recommended to discuss with others and test pilot studies (Saunders, Lewis and Thornhill, 2012).

The questionnaire is designed with three sections. The first section to ask about the basic information of participant, the most important question in this section is that “Have/had you used the e-learning system (Blackboard, Moodle, Medis, ECS, or others) of the University of Southampton?” If participant select option “No”, they will complete the questionnaire. At the second section, the questions are self-developed question to conduct descriptive statistic regarding using e-learning system of students to facilitate their KM. The content of these questions in this section is classified by attribute and behaviour. These questions based on the theory on literature review regarding e-learning system success and KM. They also have been recommended by friends who are studying PhD and have much experience with questionnaire design; and test in pilot study. The final section in the questionnaire aims to test the research framework and research hypotheses. This section includes 32 items and all adopt 5-point Likert scale (from “1 – strongly disagree” to “5 – strongly agree). All items are adapted and adopted questions used in prior researches regarding e-learning system success and e-learning system as a tool to support KM (for example, items are referred and adapted from Lin, 2007; Lin, 2007, Liaw, Chen and Huang, 2008; Liaw, Huang and Chen, 2007 ; Wang and Chiu, 2011). The scales have been tested by previous researchers as discuss in the section 2.2 and 2.4 of the literature review chapter. Thus, the reliability and validity of the instruments are high because of revealed in public papers.

Furthermore, the participant information sheet and only consent form are stated in the welcome page of the online questionnaire to introduce briefly regarding this research, researcher, research questions, contact of researcher and Ethic Committee, and a participant’s consent to taking part in the survey. Additionally, the a debriefing page is stated at the ending page of the online questionnaire in order to give thanks to participants as well as introduce briefly about this research such as research hypotheses and papers closely related to this study.

The full questionnaire is stated in Appendix 1; and the e-mails which are sent to several schools of the University and to fellow students to collect data are stated in Appendix 1; and.

3.6.3. Pilot testing and assess validity

A pilot study (pre-test) conducts a small part of sample to test the questionnaire before delivering the questionnaire to collect primary data from sample. Implementing the pilot test is crucial, especially when researchers lack of experience within designing a survey questionnaire as well as data collection approach (Yin, 2011). According to Vaus (2002), in term of pilot test, individual questionnaire items need to evaluate the variation, meaning, redundancy, scalability, not-response and acquiescent response while the whole questionnaire should test the flow, question skips, timing and interest and attention of respondents. The pilot test also aims to enhance questions in the questionnaire and it can do more than one time. Moreover, after completing the questionnaire, participants of pilot test can comment to refine the questionnaire where which questions can be misunderstood, silly or difficult (Sapsford, 2006). Additionally, the reliability and validity of items in questionnaire can be assessed with the pilot test. Factor analysis also can do in this pre-test to remove low quality items in the questionnaire (Fowler, 2008).

The pilot test to improve the questionnaire in this research was conducted two times. At the first time, the questionnaire was designed with only 2 first sections. The link of the questionnaire was sent by email and Facebook to 20 participants which are students at the University of Southampton. After that, some respondents gave much valuable feedback used to enhance the instruction as well as individual questions in the questionnaire. Because of almost self-developed questions in section 2 of the questionnaire, some questions were recommended by participants who have much experience in designing survey and statistic research. For example, in the question to ask about benefits of using e-mail function of the e-learning system, the kind of question should be changed from multiple choice questions to check box question because respondents can want to select all options. Another example is that some questions such as Yes or No questions had been improved to the open ended questions to investigate in deep why Yes or No option is selected. On the other hand, some participants who studying in English subject at faculty of Humanities at the University had contributed much valuable advice regarding the language including grammar and words choice of the instruction part as well as individual questions. Additionally, the overall appearance and organisation of the questionnaire had been commented. The time consuming to complete the two sections of the questionnaire was calculated in this first pilot test around 4 minutes. After conducting the pilot test, individual questions were changed to be clearer and easier to understand.

At the second time of the pilot testing, the completed questionnaire had been design with the final section. Because 50(+/-20) is the typical sample size at the pre-test stage, the link of the questionnaire was sent to another 40 students (Cooper and Schindler, 2010). After that, there are 32 completed answers with no missing from participants because having 02 PhD students do not use the e-learning system at the University and 6 missing answers. The alert participants if they have left any questions blank function of the iSurvey was not turned on that is the main reason of missing answers in pilot test. Therefore, this function is setup in the main data collection. From the data collection, the consistence of the multi-item scales question was tested by SPSS. This pilot data is also very useful to study data analysis in practice with SPSS and AMOS model test software. Furthermore, overall feedback from almost respondents about the questionnaire is that all questions are clear and easy to understand because the questions in the first two sections had been updated after the first pilot test and all questions in section three have been adapted and adopted from prior researches. Statistic from iSurvey administrative site, from 8 to 10 minutes is a typical time to complete the questionnaire.

3.6.4. Data collection

The completed questionnaire version, the Ethic, Risk, consent forms, the information sheet and debriefing sheet has been submitted to the Ethic committee. After approved by this committee, the questionnaire is broadcasted to students and alumni at the University of Southampton. Initially, e-mail including the questionnaire link is sent to all schools at the University in order to request a help spreading out the questionnaire to students at these schools via the e-mail system of the University (Appendix 2). Several schools (such as English, Music, Modern language at the faculty of Humanities; Law and Management school at the faculty of Business and Law; faculty of Medicine; Education, Mathematics and Social Sciences school at the faculty of Social and Human Sciences) have been agreed and helped to send the questionnaire to all their students. They replied a confirmation e-mail while some other schools sent a sorry e-mail which means cannot help to deliver the online questionnaire.

Secondly, the link of the questionnaire is shared to some Facebook fan page of the University of Southampton such as the University of Southampton Alumni, SUSU group, Southampton Management School Alumni fan page; and some Facebook group such as Vietnamese Society at the University of Southampton group; Badminton, Tennis and Table Tennis at the University of Southampton group.

The data are collected during 15 days from the 2nd of August to the 16th of August 2012.

3.6.5. Data analysis

Both descriptive statistics and inferential statistics to analyse quantitative data are used in this research. Initially, this study conducts descriptive statistics to report the averages, the dispersion, and the central tendencies of the data collected (Saunders, Lewis and Thornhill, 2012). After that, the research framework and hypotheses are tested by two-phased approach for Structural Equation Modelling (SEM) (Schumacker and Lomax, 2004; Hair et al. 2006 cited in Wang and Chiu, 2011).

In term of technology for data analysis, Microsoft Excel, IBM SPSS (Statistical Package for the Social Science) and IBM AMOS (Analysis of Moment Structures) software are used. SPSS which is well-known computer programme widely used to carry out statistical analysis in Social Science. AMOS also is a powerful tool and easy-to-use with graphical interface design to analyse model fit. Thus, using these tools can accomplish quickly the results with the highest accuracy. However, lack of basic skills in using SPSS and AMOS software can be the problem affecting the research progress. Nonetheless, due to the booming sharing knowledge in the internet, many online instruction video clips are available on Youtube and tutorials in the internet today. Thus, these tools can be controlled fundamentally in the short period of time.

The strategy to analyse data is stated below:

Analyse the descriptive statistic in SPSS

Test the reliability of items by using reliability analysis in SPSS

Analyse factor analysis to find and delete the unnecessary items in SPSS

Conduct confirmation factor analysis (CFA) to measure research framework in AMOS

Evaluate structure research framework and calculate hypotheses in AMOS.

3.7. Summary

This chapter explained about the selected research method base on the theory, literature review and the research questions and objectives. This research is appropriate with positivistic philosophy and deductive approach. Mono method with quantitative approach is suitable choice to conduct this research. Cross-sessional is the time horizon determined for this research. This chapter also discuss detail regarding data collection and analysis with choosing sampling technique, designing questionnaire, testing pilot study, collecting data strategy and analysing data strategy.