Hello, Sajid –

My short answer would have to be “no”. If the study was regionally based (e.g. in one country, or even one city), and if only one airline was sampled, results could only be generalized to that one sample/airline…that is, assuming that the sample size was adequate to detect a statistically significant effect, etc. If similar studies were conducted in different national contexts, for example, and if similar results were achieved across these contexts, then it would be more likely that results could be generalized beyond the first study to the industry as a whole, as these results would also validate the results from the first study. However, I would think that, despite certain regulatory commonalities, cultural and other differences among various countries (e.g. developed, developing, and developed nations) might produce different results. What do you think?

Ruth

Hello, Sajid –

My short answer would have to be “no”. If the study was regionally based (e.g. in one country, or even one city), and if only one airline was sampled, results could only be generalized to that one sample/airline…that is, assuming that the sample size was adequate to detect a statistically significant effect, etc. If similar studies were conducted in different national contexts, for example, and if similar results were achieved across these contexts, then it would be more likely that results could be generalized beyond the first study to the industry as a whole, as these results would also validate the results from the first study. However, I would think that, despite certain regulatory commonalities, cultural and other differences among various countries (e.g. developed, developing, and developed nations) might produce different results. What do you think?

Ruth

Hello Sajid,

I think I totally agree with Ruth, as it is not enough evidence to generalize results across countries or nations, especially if the study was only conducted in one country or even one state/region/city. As an epidemiologist I would say, though generalization of results depends on the strength of the type and sample size (representative), statistical and methodological rigor (e.g Longitudinal/Cohort, Experimental/ Ex-Post Facto Designs (Quasi)); a researcher may need to bear in mind the presence of External Environmental Factors (EEFs) and other determinants (culture, social/network groups, Unions) that differ from region to region and/or from country to country, which may affect outcome of study or the dependent variable; these factors can well be accounted for at the design stage to provide evidence-based results that can be generalized in the context and in the region/country where this study was done.

Dear Ruth and Amobi,

Many thanks for your response.

I agree that contextual/ economic differences do exist in the same industry across various countries. My study respondents are Airline flight attendants and I am using Hair et. al.`s method for sample size i.e. no. of items in questionnaire x 5. I have collected data from across the globe but approximately 90% of the data is from Pakistani airlines only.

Please correct me if I am wrong, what I understand here after going through your contributions, is that I cannot use global data to give generalizable results. Instead, I should just present my findings for specific region/country from which i have higher responses. 

Hello, Sajid -

From what you have explained thus far, I think that the best you can do is generalize results to the sample of Pakistani airlines, and note that in your report. You will have to state that results cannot be generalized beyond a Pakistani context; however, you can recommend further studies in other national contexts. 

As far as "global data" - I am assuming that you are referring to extant studies on the same phenomenon in other national contexts? If your study yields results similar to extant studies, your study can help to validate other studies. However, I would be very clear about those other studies, and pay attention to differences in results also.

 Finally, if 90% of your data is Pakistani based, I would hesitate to claim that the data is "global" in nature. It seems to me that you would be best to (a) report the regional focus on a Pakistani sample as a limitation of the study, (b) state that the sample cannot be generalized beyond the Pakistani sample, and (c) recommend that further studies be conducted in other national contexts. Does this help?

Clear answer is no since too many statistically important factors differ. Talk to a person coming from that field (I have some experience)

- short-, medium, or long-distance flight (do you usually sleep at home each night)

- recovery days after long-distance flights

- wage level

- career perspective

- which class do you work in

- quality of planes

- customer expectations in plane culturally (flying British to Mallorca is not the same as flying business persons)

- selection process of personal - if you do not pay 200% attention to personel character stability, work climate suffers

.....

Dear Sajid! I am second to Ruth, because you will have to be in need of massive sample to jstify your research, or make it a single case study in Pakistan then it could be applied.