Validity of the data is intended to obtain a level of confidence with regard to how far the truth of the results of research, reveal and clarify the data with actual facts on the ground. In qualitative research, the validity of the data is more in line with the research process as it takes place. The validity of qualitative data must be carried out since the beginning of data collection, i.e since perform data reduction, data display and conclusion or verification. To obtain the data validity in qualitative research was conducted by maintaining credibility, transferability and dependability that the intention is:
a. Internal Validity (Credibility)
Internal validity is a measure of the accuracy of data obtained with the instrument, namely whether the instrument was seriously measure the real variables. If it turns out the instrument does not measure what should be measured, the data obtained does not correspond to the truth, so that research results can not be trusted, or in other words do not qualify validity.
According to Nasution (1996: 114), internal validity (credibility) can be done by: a). Extend the period of observation, b). Conducting continuous observation, c). Data triangulation, d). Discuss with others (peer debriefing), e). Analyze the negative case, f). Using reference materials, and g). Holding member check.
In conducting this research, in order to achieve the credibility of researchers do the steps as follows:
- Extend the observation period, extend the period of observation intended to detect and take into account the distortion that might corrupt the data. Distortion can occur because of the element of intent such as lying, cheating, and pretend by the subjects, informants, key informants. Intentional element may be a mistake in asking questions, motivation, just for fun or depressing researchers.
- Continuous observation, the constant and continuous observations, researchers will be able to pay attention to something with more care, detail and depth. Continuous observation, will finally be able to find which ones need to be observed and which do not need to be observed in accordance with the data acquisition effort. Observations continuous basis to be able to answer questions about the focus of the proposed research.
- Data triangulation, data triangulation aim in this research was to verify the data by comparing the data obtained from other sources, in various phases of research in the field. Triangulation of data in this research is the source and method, meaning that researchers compare and check to return the degree of confidence the information obtained through the time and different tools in qualitative methods. The triangulation of data sources, among others done by comparing the data obtained from interviews with key informants and informant. Data triangulation was done by, first, to compare the results of the first observations with the following observations. Secondly, comparing the observed data with the results of the interview. Comparing data from the first interview with the results of the next interview. The emphasis of the results of this comparison is not a problem in common ideas, opinions, thoughts alone. But more important is to know the reasons for the differences.
- Discuss with others (peer debriefing), discuss the results of the data with others who are familiar with the research that is being done.
- Analyze the negative case, analyzing the negative case the intention is to seek the truth of the data that is said to be true by a data source but rejected by other sources.
- Using reference materials for comparison and to sharpen the analysis of data.
- Holding member check. The purpose of holding member check is that the information has been obtained and will be used in the writing of the report can be in accordance with what is meant by the informant, and key informants. Therefore in this study member check carried out each end of the interview by repeating outline answers or views as researchers record the data based on what has been said by the respondent. This objective is carried out so that the respondent can fix what is not appropriate according to them, reducing or adding what is lacking. Members check in this research conducted during the study-as formal and informal interviews run.
b. External validity (transferability)
External validity with respect to the problem of generalization, i.e to where generalizations are formulated also applies to cases other than research. In qualitative research, the researcher can not guarantee the validity of the results of research on other subjects. This is because qualitative research does not aim to generalize, because in qualitative research does not use random sampling, or always be purposive sampling.
Dependability or reliability of the instrument is an index indicating the extent to which gauges trustworthy or reliable. Reliability indicates the extent to which the measurement results remained consistent when done again against the same symptoms with the same gauge. To be able to achieve the level of reliability in this study, it is done with the technique of repeated or check recheck.
In qualitative research the researcher should try as much as possible minimize subjectivity factor. Research will be said to be objective when warranted or "confirm" by other researchers. So objectivity is identified with the term "confirmability".
That all about the Validity of Data for Research Instruments on Qualitative Research. Understanding more about qualitative research it will help the researcher get more unique things in her/his research.
Hopefully helpful. Thanks