Dependability in Qualitative Research

Dependability in Qualitative ResearchWe can define dependability in qualitative research as the stability of data over time and over conditions. Dependability can be compared to reliability in quantitative studies. In other words, dependability is an evaluation of the quality of the integrated processes of data collection, data analysis, and theory generation.

Dependability could be looked at as the researcher’s account of the changes built into any setting in addition to changes to the research design as learning unfolded. Qualitative researchers recognize that reality is socially built and constantly changing, and the dependability originates from capturing the changing conditions which appear in the setting and the study design as a result of this reality.

As stated by Lincoln and Guba, a dependable study needs to be accurate and consistent. Two ways of evaluating dependability of data include stepwise replication and inquiry audit.

Stepwise replication entails a number of researchers who can be divided into a pair of teams to carry out separate inquiries with a view to comparing data in addition to findings. Inquiry audit means data  and relevant supporting documents being scrutinised by an external reviewer.

To be able to deal with the dependability issue more directly, the processes within the research needs to be documented at length, thus allowing a future investigator to repeat the work, if not necessarily to obtain the identical results. Thus, the research design may be considered a prototype model. This kind of in-depth coverage also permits the reader to assess the extent to which proper research practices are actually adopted.

In order to allow readers of the research report to develop a comprehensive understanding of the methods and their effectiveness, the text needs to include sections devoted to:

  • the research design as well as its implementation, explaining what was planned and carried out on a strategic level;
  • the operational aspect of data gathering, addressing the details of what was completed in the field;
  • reflective assessment of the project, evaluating the potency of the process of inquiry carried out.

Ways to Improve Dependability in Qualitative Research

Dependability is associated with the consistency of findings. Since many qualitative methods are customized to the study situation, the specific ways of data gathering, analysis, and interpretation in qualitative research should be described. This kind of thick description of methods offers information as to how repeatable the research could be or how unique the situation is.

Step wise Replication Approach

This approach should be included in the design of a qualitative research to improve dependability. Two research teams deal individually with data which have been split, and the results are contrasted.

A crucial factor in undertaking stepwise replication is that communication among teams and team members is vital. There should be a provision for communication on a regular basis and at preset points in the research process.

Code-recode Procedure

A researcher should conduct a code-recode procedure on his data throughout the analysis phase of the study. After coding a section of data, the investigator needs to wait at least two weeks and then return and recode the same data and evaluate the results.


Dependability could also be improved by means of triangulation to make sure that the weak points of one approach to data collection are reimbursed by the use of alternate data-gathering methods.

You can improve stability over time by repetitive observation of the same event and re-questioning people about key issues

The use of fellow workers and methodological experts to examine the research plan and execution is yet another way of guaranteeing dependability in qualitative research.


  1. H. Howard says

    The US government should use qualitative research when, and after they implement programs. It would save tax payers a lot of money. In old school terms, this is call checks and balance. You can balance at a average, a 60%, 70%, or peak level depending on your performance expectations. Code-recode is nothing new, they are just involving more people, businesses, and ethics. Be careless what you ask for, the mine is biological, and changes with time. Correct data analyzing is as important as collected the data.

Speak Your Mind