Social science research methods (2)

Social science research methodsHuang Ying School of Foreign Studies, CUMT

1.5 Ethics in social science researchThe researcher should be ethical and adhere to basic principles of ethics of social science research. “Scientific misconduct occurs when a researcher falsifies or distorts the data or the methods of data collection, or plagiarizes the work of others” (Neuman, 2007, p. 49).

“Research fraud occurs when a researcher fakes or invents data that he or she did not really collect, or fails to honestly and fully report how he or she conducted a study.” (ibid., p. 49)

Some of the basic principles of ethical social science research internationally acknowledged so far are listed as follows (ibid., p. 60):

Ethical responsibility rests with the individual researcher. Do not exploit subjects or students for personal gain. Some form of informed consent is highly recommended or required. Honor all guarantees of privacy, confidentiality, and anonymity.

Do not coerce or humiliate subjects. Use deception only if needed, and always accompany it with debriefing. Use the research method that is appropriate to a topic. Detect and remove undesirable consequences to research subjects.

Anticipate repercussions of the research or publication of results. Identify the sponsor who funded the research. Cooperate with host nations when doing comparative research. Release the details of the study design with the results.

Make interpretations of results consistent with the data. Use high methodological standards and strive for accuracy. Do not conduct secret research.

1.6 Reliability and validityReliability: the consistency and replicability of research, comprising two subsub-types: internal reliability (e.g. interinterrater reliability) and external reliability.

Internal reliability refers to the consistency of data collection, analysis, and interpretation. External reliability refers to the extent to which independent researchers can reproduce a study and obtain results similar to those obtained in the original study.

Validity: has to do with the extent to which a piece of research actually investigates what the researcher purports to investigate, which includes two types: internal validity and external validity.

In addition to internal and external validity, a third important validity is construct validity, which has to do with the question: Is the study actually investigating what it is supposed to be investigating?

Internal validity: the interpretability of research. Question: Can any differences found in experimental research be ascribed to the treatments? External validity: the generalizability of results from samples to populations.

Table 1 Questions for establishing the reliability and validity of a study

(Nunan, 1993: p.17)

1.7 Types of dataNote that differences in data types (or variables) or measurement scales determines the application of statistical methods in research. Qualitative data (variables): nominal and ordina

l. ordinal. Quantitative data (variables): metric (or interval).

Nominal data (名义数据): names or (名义数据): categories of phenomena which cannot be treated numerically in terms of such operations as computing averages. Ordinal data (顺序数据): those can be ordered (顺序数据): or ranked according to some hierarchical system such as test scores, degree of presence of some characteristics, or the relative number of occurrences of some sentence type among a group of subjects.

Interval data (定距数据): those that have (定距数据): numerical values and that can be manipulated mathematically. “The intervals between points are of equal value over its whole range” (Argyrous, 2005: 10). Ratio data (比率数据): those that have a (比率数据): true zero point over the whole range such as age in years, etc.

Note: data types are changeable one way or another; we categorize nominal and ordinal data/variables into categorical/discrete variables and metric or intervals variables into continuous variables.

Cited from Statistics for research with a guide to SPSS, p. 11.