• Nenhum resultado encontrado

Validity, reliability and minimizing risks

Validity and reliability are terms linked with the trustworthiness of a research. Validity measures how accurately the data collection methods measure, what they are supposed to measure and to which extend the findings are really about what they profess to be about (Saunder, Lewis & Thornhill 2007, 614). With reliability it is referred to which extend the data collection techniques will produce consistent findings, if other researchers would make similar observations and conclusions or if there is transparency in how sense was made from the raw data (Saunder, Lewis & al 2007, 609). These are linked to the discus- sion about forms of bias within qualitative research. As this research has mainly been qualitative, this chapter is concentrated on bias, risks and requirements related to qualita- tive research and its analysis.

The validity and reliability are harder to measure in qualitative research than in quantita- tive research because in the qualitative research the researcher plays a crucial role in the whole research process. Especially the analysis can be sensitive to the risks related to validity and reliability of a research (Eskola & Suoranta 2001, 210) because several inter- pretations must be done. First, the interviewee interprets a phenomenon and this is fol- lowed by interpretations made by the researcher about the material and then when writing a report which is then interpreted once more by the reader (Eskola & Suoranta 2001, 141). The report normally includes only the most essential information understood by the researcher (Eskola & Suoranta 2001, 141) and that is why there is a risk of missing some

essential piece of information from the interpretations. (Eskola & Suoranta 2001, 141, 208-213.)

The interviewer can easily affect the research reliability. The interviewer’s comments, tone or non-verbal behaviour may have an effect on the responses he or she gets. This is called interviewer bias. In this paper the researcher tried to minimize the risk of interviewer bias for the reliability of the paper in different ways. First, it is important to understand as- sumptions one might have. Before running the semi-structured interviews the researcher understood that she assumes that different areas of HRM can and should be developed to be done better. These assumptions were based on the earlier phases of the research and the studies in HRM. The philosophy of HRM the researcher may affect the set of ques- tions and from which approach the situation is approached. This risk was reduced by re- cording all the interviews and reviewing them during the analysis phase so that the re- searcher actually listened carefully what the interviewee said, not only assuming what an interviewee meant with it.

Reliability is linked with the credibility of a research (Eskola & Suoranta 2001, 211). This means that the researcher must make sure that the interpretations done by the researcher are similar with the respondent’s interpretations. I therefore asked during the interviews more detailed questions and sometimes even the same question in different words about the same topic in order to make sure that I had understood the respondent’s statement right. Most of the time the interviews agreed with my interpretations but once they ex- plained their idea from the different point of view.

Besides taking notes and records and repeating questions the interviewees, the inter- viewer bias was reduced by careful planning and preparations. The level of knowledge of the participants was thought beforehand as well as level of information supplied to the interviewees, appropriateness of location, attentive listening skills, the scope to test un- derstanding and approach to recording data. The interviewees were chosen to be repre- sentative part of HRM professionals working for the company while having a good knowledge about the local HRM practices. In a case of not being familiar with the termi- nology used in the interviews the interviewees were provided a list of definitions well be- fore the interview took place. The interviews were conducted face-to-face, and the outfit was the casual office outfit as should be when working at the headquarters, and through Microsoft Lync because the interviewees were abroad and then conversations could have been recorded within the program. Attentive listening skills were shown by asking more detailed questions and making sure I had understood their statements right by summariz- ing what they had said. Also the face-to-face interviews were recorded. This planning also

supported the validity of the research as it prevented from major mistakes of the topic of the interview for example turning to subject irrelevant for the research and that appropriate data was collected. Also possible culture differences and especially language challenges were taken into consideration and the research pattern with definitions was sent to the interviewees beforehand and the mother tongue of the respondents was used if possible in order increase the understanding between the respondent and interviewer. During the focus groups the participants had also definitions next to the questions in order to make sure that everybody was answering to the same question.

During the analysis the interviewer bias was reduced by doing analysis in little parts and representing the results in a way that the reader can also analyse their validity them- selves. The researcher also discussed about the topic and the results with her colleagues in HRM, both academic and practicing, asking their opinions and she also discussed with the CEO and other people about the HR processes.

Also interviewee might be sensitive to bias. Therefore, the risk of not getting truthful an- swers or the respondents not answering freely: the risk reduced by creating a relaxed at- mosphere assuring that the sensitive information would be reflected in a neutral and anonymously in the report.

There have been challenges which can affect to the validity of the research. Firstly, the topic is very wide and it must have been therefore demarcated very tightly. If seeing only the research question, one could assume that the findings don’t cover only the develop- ment of HR processes but also the structure of HRM. However, HR processes are only part of a bigger HR architecture as explained in Chapter 2 and then if the whole HR archi- tecture isn’t functioning properly, neither the HR processes can. When understanding the theory behind the HR processes also the findings are considered to be valid especially in this case study. Also other research methods can be seen as valid and logical approach to the case company’s situation. The research aimed to develop the HR processes at group level. The group level has a strategic importance at the geographical level too, as it is supposed to coordinate operations in all group’s geographical locations. The geograph- ical coverage was considered when designing the research as the focus group partici- pants came from the most important areas of HRM covering three, four continents. Also the interviewees were from different countries covering how the roles of different-sized units often vary.

4 Findings

In this chapter the results of the research are presented investigative question by investi- gative question. The data was analysed according to the research design. First, the cur- rent HR processes are discussed research phase after each other and some trends are concluded in the end. Investigative questions two and three are analysed similarly. Inves- tigative question four analyses the themes of the first three investigative questions from the development point of view based on the theory. It also shares other development sug- gestions for HRM process of the case company.