Expert judgment is the primary method used to determine whether a test has content validity. In factor analysis, convergent validity is tested by assuring  that all items yielded significant loadings, and at the same time, all the item-total correlations were above .50, providing evidence of high convergent validity.

Dear @Ratan, Choice of test: In terms of selecting a statistical test, the most important question is "what is the main study hypothesis?" In some cases there is no hypothesis; the investigator just wants to "see what is there". For example, in a prevalence study there is no hypothesis to test, and the size of the study is determined by how accurately the investigator wants to determine the prevalence. If there is no hypothesis, then there is no statistical test. It is important to decide a priori which hypotheses are confirmatory (that is, are testing some presupposed relationship), and which are exploratory (are suggested by the data). No single study can support a whole series of hypotheses. See the full text at:

http://www.bmj.com/about-bmj/resources-readers/publications/statistics-square-one/13-study-design-and-choosing-statisti

Dear @Ratan, my major is not statistics, maybe the experience and the answers of many researchers of the following thread could help you! I do think so!

https://www.researchgate.net/post/Should_we_re-test_the_scale_reliability_and_validity_of_a_good_questionnaire_in_order_to_perform_a_study_in_multicultural_cross-cultural_context

Expert judgment is the primary method used to determine whether a test has content validity. In factor analysis, convergent validity is tested by assuring  that all items yielded significant loadings, and at the same time, all the item-total correlations were above .50, providing evidence of high convergent validity.

Dear @ Abedallah, thanks for such a good suggestion...

Dear @ Mahafuz, thanks for suggesting factor analysis. Actually I want to clear some more from you that what are validity components I am supposed to test for scale translation? 

Dear @ Ljubomir, Kamal and Shafig, thanks for the inputs...

Hi Ratan,

It's a bit tricky to comment in detail without more details from your study, but Cronbach's alpha statistic might be a useful measure of internal consistency in you case:

http://en.wikipedia.org/wiki/Cronbach's_alpha

On the other hand, whether it is useful will depend on how you constructed your questionnaire and the complexity of the underlying variable you are trying to assess. In the case of convergent construct validity, I would say that your choice of approach will be highly dependent on the availability of data from other scales to compare with your own.  Rather than focusing too quickly on a specific test as the one and only answer, I think it might be useful to think about what the key questions are that you want to address and then think of the process that will be needed to answer those questions. In that way the appropriate test statistic might suggest itself.  Again depending on the availability of comparative data, some kind of data resampling, empirical, test rather than a parametric test might turn out to be most appropriate.

Dear @ Nail, I agree with you and thank you so much for your valuable statements...

If you have develop already some scale measures and you want to apply to some multicultural research, one should continuously make tests of goodness of fit of the given scale to be applied to a particular case. After you decide the model then you proceed as ordinary. 

Ratan, I think it would be advisable for you to ponder on the "wisdom" of Abedallah's observation concurred by Neil's. You must distinguish between "measuring" validity or calculating the magnitude of the validity, and testing if such magnitude in the sample also exists in the population. Tests, in statistics are always inferential.

·         Dear Ratan Khuman

It depends on the scale development or inventory buildup. If the scale is new then factor analysis then construct validity should be applied. If the inventory is available, then how are the factor structure, discriminant validity and the internal consistency of the scale should be test it?

Hi Ratan,

You may find this article on statistical estimation of internal validity useful:

http://www.personal.psu.edu/ljk20/randtests.pdf

Dear @ Costas, Eddie & Hashem, I agree with you and thanks for valuable suggestion...

Dear @ Monica, thanks for the interesting link...

Dear @Monica, this is a very interesting article; it deserves reading and discussion.

Thank you, Abedallah. Appreciate it.

Dear Clayon, I agree with you and thanks for the input...

This depends on the validity in question. There are several types of validity: concurrent validity, predictive validity, construct validity, and face validity. Each of these types of validity requires a different type of test. Nevertheless, the first test of any validity is whether or not you are measuring exactly what it is that you claim to be measuring. Do you measure exactly what it is that you say you are measuring?

Using a scale specifically designed to weigh only blue whales, for example, would not be a valid measure if you are using a blue whale scale to measure the weight of a mouse.

Dear Ratan,

Recently, I worked on a very similar paper about the construct, convergent and predictive validity of a questionnaire consisting of 33 four-point Likert-type items. In the paper you may find useful information about the exploratory factor analysis, confirmatory factor analysis, and other procedures used for testing the validity of non-normally distributed ordinal-scale variables.

I hope this helps.

The reference is:

Pedisic, Z., Matouschek, S., Stuckler, M., Basaric, A., Titze, S. (2014). Reliability and validity of the German language version of Girls’ Disinclination for Physical Activity Scale. European Journal of Sport Science, DOI: 10.1080/17461391.2014.899397

You can download the article here:

http://www.tandfonline.com/eprint/EBhvNtAfCvYid7rZhNDB/full

Dear @ Zeljko, thank you so much for the valuable attachment...

A good book to read on this subject is "Quality of life" by Fayers and Machin. It goes into great detail on the subject of validity for questionnaires.

Internal validity is not a property of questionnaires, but has meaning for scientific studies. Convergent construct validity is basically an attempt to show that a questionnaire has construct validity.

At this point I suggest you to tell us your hypothesis like Abedallah said. What will you try to find in this study? This will make it easier for others to help you choose a statistical test. But don't go looking for a statistical test first. Your first concern should be your hypothesis and then the design of the experiment you need to do.

Hi Ratan

you might find this book useful: Streiner DL, Norman GR. Health measurement scales: a practical guide to their development and use. Oxford university press.

Regards

Pieter

I used rasch analysis but you could use crown aches alpha for internal validity of a scale or item response theory. Rasch can convert to a continuous scale score so it is good from the hat view point. I also used correlations for convergent and divergent validity, easy to do and explan!

regards Linda