• No results found

Reliability and Validity ···································································

5.4 Data Collection ··················································································

5.4.3 Reliability and Validity ···································································

The reliability and validity of the data obtained in this thesis is strengthened because it

has met the qualitative criteria of credibility, transferability, dependability, and

confirmability suggested by Lincoln and Guba (1985).

Credibility

First, as the researcher of this study, I can understand the phenomena of interest from the respondents’ original data. Secondly, although many subjective views emerged from the interaction with respondents, I always avoided imposing my own

understanding to their responses. The credibility was achieved by no less than two

hours prolonged and sustained engagement in each long interview or focus group

Transferability

The results are not only transferable within the population within this study but also

applicable beyond the bounds such as to the Chinese university students from other

cities that plan to go to the U.K for higher education. This is because Chinese students

from different universities have many similar attributes and the same shared Chinese

culture.

Dependability

The interviews were audio recorded which has assured the thoroughness of data

documentation. Therefore, the data collected is stable and consistent over time. In

addition, as discussed in the previous section, the refined McCracken (1988)’s five

stages of analysis advocated by Herrington and Oliver (2000) was used because of its

high dependability.

Confirmability

Triangulation can act as an important technique for establishing confirmability

(Lincoln and Guba, 1985). The methodology of the long interview and focus group

interview drawing from the same paradigm were used to triangulate, which can

effectively reduce bias in data sources.

In addition, this thesis has gone through two additional validity procedures within the

qualitative lens and paradigm assumptions suggested by Creswell and Miller (2000).

These three validity procedures include the audit trail, researcher reflexivity, and thick

description. First of all, the example data matrices as shown in Appendix 9 provide

clear audit trail. In addition, the study’s internal validity is also enhanced through the

use of a questionnaire. Secondly, the researcher reflexivity was achieved by my

review of cultural categories which is illustrated in Chapter Three. Thirdly, thick

of acquiring external validity.

5.5 Conclusion

This chapter has explained the aspects of the research philosophy of this thesis, the

four-stage method suggested by McCracken (1989), and how the data was collected.

Following this, the next chapter will analyse the data collected. It refers to stage four

of the method discovering analytic categories and analysis write up (McCracken

1989). It will include the data analysis of the 23 long interviews and the data analysis

of the three focus group interviews. The difference between the findings from the long

Chapter Six: Data Findings and Analysis

6.1 Introduction

The long interviews triangulated with focus group interviews can result in a wealth of

knowledge which is less likely to emerge from an impersonal questionnaire. One of

the primary reasons for the use of long interviews and focus group interviews rather

than questionnaires is to explore a diverse set of responses. According to McCracken

(1989), this chapter refers to Stage Four, discovery of cultural categories and analysis

(Figure 6.1).

Review Process Discovery Process

Figure 6.1: Stage 4 Discovery of Analytic Categories and Analysis

The sample data matrices of the long interviews are available in Appendix 9. Symbols

were used for reducing large amounts of data into a smaller number of analytical units.

There were 85 influences identified that can influence the five steps as shown in

Appendix 7. In order to reduce large amount of data into smaller number of analytical

units, symbols of (1), (2), (3), and (4) were used and written after each influence

identified, thus indicating the four steps in their decision-making process as suggested

in Figure 6.2. Examples are shown in the table 6.1.

Stage 1

Review of analytic categories and interview design

Stage 2

Review of cultural categories and interview design

Stage 3

Discovery of cultural categories and interview Stage 4

Discovery of analytic categories and analysis/ write up

Cultural Data Analytic Data

Table 6.1: Examples of Coding by Symbols

Transcript Coding by symbols Expanded meaning

“Because I think to learn a language is a long-term career. Four years [of study in the university] is not enough. It’s far from enough. If I really like this career, I would spend a lot of time on it” From UK1 (p.2)

Knowledge acquisition (1) The influence of knowledge acquisition can influence the respondent’s Step One (deciding to continue to study after graduation)

“Generally speaking, the academic reputation of overseas universities is much better than Chinese universities. This is an important reason I decided to go abroad for further study” from UK1 (p.27)

Better academic reputation in overseas university (2)

From UK1 (p.27)

Better academic reputation in overseas university is a motive that can influence Step Two (deciding to study abroad)

“L: Have you ever considered Singapore as your study destination because they speak English?

R: No, it is an Asian Country, and Chinese dominant country. I hope I can go far away from China. I have a strong curiosity about a completely different country and the U.K is a good choice”. from UK1(p.30.

Different cultural environment (3)

From UK1 (p. 30)

The influence of different cultural environment can influence the Step Three (selecting a host country), which means the respondent wants to choose a host country that has perceived different culture from China.

In addition, abbreviations were also used in the data analysis. Examples are shown in

the table 6.2. The data was sorted in order to provide the basis for the last stage of

writing up.

Table 6.2: Examples of Coding by Abbreviations

Transcript Codes by Abbreviation Meaning

“It is said that the air in the U.K is fresh and the roads are clean. There might not be many skyscrapers but old red-brick buildings are more attractive and good looking.” From UK2 (p5).

GCA-UK (pos.)

Good living

environment

Good living environment is one of the positive General Country Attributes of the U.K

“The first impression of British services is the high price because the labour cost is high in the U.K.” from UK2 (p.6)

GSA-UK (neg.) Expensive

British services are expensive is one of the negative general service attributes of the U.K

“The most important reason is due to the expansion of enrolment. Too many low quality students can study in the universities. The overall quality of university students is decreasing.”

From UK3 (p.13)

SSA-China (neg.) Enrolment expansion

The enrolment expansion is one of the negative attributes of Chinese higher education services

“In China, some students are likely to cheat in their assignments and exams, not all students. However, those who do not want to cheat may change their mind to cheat because many students cheat without punishment. Students do not worry about cheating, and then they tend to be lazy. In the U.K, academic cheating takes risk. Those who do not want to cheat will be confirmed to be hard-studying. Those who are likely to cheat will change their mind to be hard-studying because of the possible punishment of academic cheating. In my opinion, British higher education can produce more quality students than Chinese higher education.” From UK3

(p.14-15)

British SSA > Chinese SSA

Academic misconduct

The attributes of British higher education are preferred to those of Chinese higher education in terms of beating academic misconduct