empirical study is always based on experiments and experience which support infallible knowledge rather than ideas and theories.

You can extend the conclusions of your study if the interpretation of the results was reinforced and supported by statistics and statistical evidences, and then you will be able to generalize your results.

Thank you for the pertinence of your response.

empirical study is always based on experiments and experience which support infallible knowledge rather than ideas and theories.

Thank you for the clear answer I appreciate a lot.

Hi Adriana,

Yes. you can do so but since qualitative research is based upon a different epistemological paradigm than quantitative research (constructivism rather than positivism) it also has (somewhat) different methodological assumptions. Predictive statistics are not employed the same way (if at all) and the type(s) of generalization is different. Skipping the idea of internal vs. external validity here for expediency (though the interaction between the two is an important consideration), in experimental research generalizability refers to the extent to which the conclusions in your study would hold outside your de-contextualized experimental design. That is, external validity is related to generalizing. There are two major approaches to how we provide evidence for a generalization experimentally : The Sampling Model which focuses on how representative your sample is of the population from which it was drawn so that you can generalize from a small sample to the population. This is where predictive statistics come into play. The “predictive” refers to predicting the small-to-many. The Proximal Similarity Model focuses on generalizing to different contexts (outside of your experimental context) and developing a theory about which contexts are more like your study and which are less so. To the extent that you can demonstrate how similar or different your contexts, people and time are to others, the potentially better you can be at conceptualizing how well you can generalize your results. This is not based on predictive statistics but, rather, on some implicit theorizing. There is even a suggestion for creating and defending a gradient of similarity. This a major way that qualitative research generalizes as well.

Generalization in Qualitative Research

There are at least two ways that generalization is applied in the qualitative research paradigm: Analytic generalization- This approach has a different set of foundational beliefs that do not rely on sampling and probability theory. Instead, operating from the process theory of human behavior and trying to account for the stratified complexity of reality, the significance and utility of the research findings are based on the awareness and operation of underlying principles/mechanisms that give rise to the social events. Within the research activities, the social events are investigated but the qualitative analyses are directed to the underlying entities and structures from which these events emerge. These underlying structures, then, are seen as potentially unifying foundational entities that may be operating across many or all participants depending on other contextual variables (Bryman, 1988; Silverman, 2000).

Consequently, in qualitative research the application of findings is not determined by the mathematical probabilities regarding the recurrence of behavior but, rather, by an understanding of the underlying entities, mechanisms and structures from which human social action emerges.

Second, Case-to-case transfer – Occurs whenever a person in one setting considers adopting a program, finding, or idea from another setting.

§ This form of generalization has a long history of utilization and this practice of transferring learning and findings from one case to another has occurred both within research settings and in areas like law and medicine.

§ Case-to-case transfer is different from the other two types of generalization in that the decision to transfer or generalize is made by the reader of the research and not the researcher.

§ For this to occur, however, the researcher has an obligation to provide a rich description of the case. (Firestone, 1993). This is where scientific realism is beneficial. In effect, it provides an agenda for the researcher in terms of description because it directs a focus on the three types of data (experience, events entities) with a particular focus on the underlying entities. As House suggests,

§ When employing this form of research generalization, it is important for the qualitative researcher to know the essential operational conditions in the first instance (case one) and to determine whether they also operate in the second instance (case two).

I have attached more detailed notes and an article that might help. Hope this was not too long!

Jack

Thanks for the extensive and consistent specifications. I appreciate the support and details of the text and the materials offered.

The establishment of truth is one of the main objectives in empirical research . More specifically, the task of an empirical researcher is to find out whether the generalizations introduced by the theorists are true. The main reason is that theory and practice are two sides of the same coin enjoying a symbiotic relationship. This means that neither can thrive without the cooperation of the other.

I appreciate and thank for the depth and clarity of the answer

generalization is dependent on sampling kind and size . data should be triangulated by different methods, theories and investigators to be generalizable . the data should be froma representative sample of population to be considered as a general scientific fact of great application

Sampling is a problem requiring much scientific rigor to allow extrapolation later. Thank you very much for the answer.

we can not directly say yes to this extent we can generalize our results. I mean the extent of generalization follows a pattern of regularity that repeats over different circumstances and that can be described simply by mathematical, graphical, or symbolic methods .as mentioned if our Sampling will be satisfying we can adjust this extension.

I appreciate the answer and thank you!