I always recommend first establishing the reliability and validity of a new measure in a pilot study. You don't want to put in a lot of time and effort collecting data just to find out (after the fact) that your measure is no good! However, you mentioned that you took questions from a previous questionnaire (which has been validated in previous studies, I presume) and included some new questions. What is the ratio of old to new questions? For instance, if the questionnaire has 20 items, with 18 old and 2 new questions you created, I don't think a separate study is needed. However, if you have, say, a 50-50 split between old and new questions, I would collect some pilot data first to make sure your measure is reliable/valid before moving forward.

Publishers and some journals will have regulations. Even places like Facebook has some rules. But there is no global laws on what you can type up. If you had asked should you do something, the answer of course would be different, but you said "mandatory," "necessary," and "required."

I Have the same question If you have got the answer please let me know.

Payam Mohammadi Sure, of course!

I always recommend first establishing the reliability and validity of a new measure in a pilot study. You don't want to put in a lot of time and effort collecting data just to find out (after the fact) that your measure is no good! However, you mentioned that you took questions from a previous questionnaire (which has been validated in previous studies, I presume) and included some new questions. What is the ratio of old to new questions? For instance, if the questionnaire has 20 items, with 18 old and 2 new questions you created, I don't think a separate study is needed. However, if you have, say, a 50-50 split between old and new questions, I would collect some pilot data first to make sure your measure is reliable/valid before moving forward.

I think the answer "Yes"

The answer is, it depends. If you are using APA, they have a section titled JARS which will give you your answer.

The content validity of a new questionnaire should be tested by a group of experts to show whether the questions are related to the subject and the objectives or not. I think you have used the Likert scale, and you can calculate as you said the Cronbach's alpha for the reliability of the questionnaire. I think for the new questionnaire, you have to write about the validity and reliability. Many journals don't stress on this point, but I think nursing journals give it a great attention.

Thank you, if the questionnaire had been previously validated for the same category including ethnicity, I think no need for a second validation.

Atira,

Many researchers gather data on bulk questionnaire (200-400 questions) and conduct factor analysis to determine which questions are associated with the factors which constitute the construct (e.g., Billard's Attitudes Towards Transgender Individuals provides an excellent example of the process for determining the factors constituting the construct in question and how to establish validity). As you mentioned, this would be during the pilot phase of your test development. Cronbach's alpha measures how related each question is to one another and is essentially the mean correlation across questions. But what you need to do is determine the factors. Otherwise, you are saying "yes, all these questions are related, but we don't know why or if some or more related than other and vice versa." Reference Billard's method for determining the significant factors in the article listed below.

Attitudes Toward Transgender Men and Women: Development and Validation of a New Measure

📷Thomas J Billard*

I hope this helps!

Bests,

Jordan Spengeman

If the international organizations of evaluation and measurement of test recommend to report the evidences of validity and reliability.

AERA, APA, NCME. (2014). Standards for Educational and Psychological Testing. Washington, DC: American Educational Research Association, American Psychological Association, National Council on Measurement in Education. https://www.apa.org/science/programs/testing/standards.

Validityand reliability is very key to a newly developed tool to ensure that the all the research objectives are answered and the data collected are representative of the whole study area

Edward Christopher Yo, please forgive me if I address only some of the issues that your question gives rise to. First, I think it is definitely a good idea to try to check out both content and face validity - both of which I suspect you've already done. That's because many scales don't transfer well to new linguistic and cultural contexts.

After that, there are a number of options for checking reliability and validity. I recommend looking carefully at interitem correlations as one form of reliability (internal consistency), and, much as they are often cited, I wouldn't lay a lot of store by coefficient alpha values. They are too often of little value and they are also widely misunderstood.

It may pay to look at the factor structure of your questionnaire to see which items might be combined to form domains. The authors of the article you provided the reference for have attempted to do that, but I have a number of concerns regarding their research. For example, I'm pretty sure their sample size is too small for some of their analyses, they use the Kaiser criterion for determining the number of factors in their data despite that criterion having been criticised for more than 30 years, and they use varimax rotation, which has also been criticised (for at least 2 decades). They also don't seem to appreciate that there is a difference between exploratory factor analysis and principal components analysis.

So, although you might want to use something such as exploratory factor analysis to check on the FACTORIAL validity of your scale(s), I would not use the article you cited as an exemplar.

All the best with your research.

Dear Edward Christopher Yo ,

You certainly have received some attention with your interesting question! I suppose one answer would be what might happen if you didn't undertake these checks. Firstly, without a validity check you have no confidence that your items actually measure what you think they do. Secondly, without a reliability check you don't know whether a group of items can be combined to measure the same construct. Validity could be checked informally but it is very important. Reliability checking could be omitted if you are only interested in analysing individual item responses.

It depends on the source and percentage of questions being used frim the existing questionnaire. If questions have been contextualised without major changes, mentioning it in your methodology suffice. However, if you have many new qyestions, validity and reliability tests are important to make your findings credible and acceptable. Use expert views, piloting and statistical test to show the questionnaire is reliable and valid

Yes, very essential to pretest your new ly developed instrument inorder to ascertain the reliability.