I am currently working on developing a conceptual framework in the construction management field. I would like to know about different methods for empirically validating such a framework.
Many thanks
Good question.
Any book on research methodology answers this in detail and you should really read at least one methodology book before you start your research.
The best way to validate a conceptual model is to do good empirical research with your framework and competing frameworks to see which fit reality better. Structural Equation Modeling is great for this if your data fits. This is known as the scientific method.
Now, if you mean, "Before doing research, how can I my thesis committee that my conceptual model is reasonably good"? the answer is different.
The best way is to ground your model in empirical literature, then you may use meta-analyses or statistical validity measures to demonstrate the validity of your framework and even of the measures you plan on using.
However, if your framework is based on your own original thought and analysis, you will have to explain the logic behind each step of its development. This is extremely dangerous for a new researcher.
Best regards.
John
Thanks John,
Your directions are very helpful. I really appreciate you advice.
Best Wishes,
Hassan
I agree with John. Conceptual frameworks require a stringent process of validation. The sample size has to be large (refer to Tachbacnick: Multivariate Analysis). You need to determine the scale that you intend to analyze too. If you are using a dichotomous scale (Yes/No or Likert Scale), you will need other programs other than SPSS (STATA/R Statistics) to perform polychoric/tetrachoric correlations.
You need to be fluent in factor analysis, validity and reliability assessment. Most importantly, you need to pilot your questionnaire well before even attempting to perform formal data collection.
Hope this helps.
Rabin
Yogarabindranath : Actually, SPSS offers logistic regression, multinomial logit and a host of techniques for analysing nominal dependent variables.
There are some good links to resources that might help answer this question here: https://www.researchgate.net/post/What_is_the_procedures_of_framework_validation_in_qualitative_research
please refer to the following link:
Article Validity, reliability, and generalizability in qualitative research
One of the best way of validating a conceptual framework is to go back to the participants to investigate if your model capture their thinking. Any feedback you obtain can be used to emend the framework.
- Badges
- Science topic
- Similar topics
- Engineering
- Modeling