You have used both qualitative and quantitative research techniques. The interviews and creation of themes (attributes) is a separate study in itself. Refinement of patterns are essentially qualitative and done intensively (rich thick description of the data collected) can be written as such (no hypothesis needed). The creation of a Likert-scale instrument on the other hand is a separate quantitative effort where a hypothesis would be typically used. As you have explained, a lot of work is yet to be done. Validity and reliability efforts are needed, e.g. Jury of Experts to review your instrument for validity (content, face, elimination of negatively written items, etc.). Then there should be a pilot of 10-15 of like respondents (similar to those that will complete the survey/questionnaire). That pilot would allow you to work through reliability, e.g. scale reliability (Cronbach's alpha, etc.). You will be able to eliminate bad items and create constructs (concentrate like items) to match your patterns discerned in above qualitative analysis. These are just the basics. Others on this forum can assist with even further detailed questions. Hope this helps. Good Luck! Regards, Professor C

Thanks a lot for your comment Prof John W. Collins. All measurement of reliability, validity done . My questions what other statistical measurement should i follow and achieve to complete the measurement of perceptions leadership effectiveness.

Your instrument/questionnaire can be piloted with ca. 20 faculty members, but a larger group should be used before you try to test any hypotheses. You should have developed some themes from your interviews. You can place/group your instrument items (questions) together to create constructs. Then you can analyze using correlations. In fact you could build the constructs using Cronbach's alpha (in SPSS, etc.) to weed out any bad items/questions. The complete construction of your instrument/questionnaire is worthy to write-up in and of itself. Then you will have Likert scale Means/Averages for each and can compare those results to a wider group. That allows all kinds of comparisons of means, e.g. t-tests, Regression etc. You will need a statistician to help you work through the weeds. The t-tests have inherent hypotheses as you compare your means to a wider group. Hope these clarifications help. Regards, Dr. C

Thanks again Dr. Collins.

I constructed my questionnaire based on the themes achieved from interviews. I tested the instrument using Cronbach's alpha, and I collected the data from 180 staff. As it is an exploratory study based on qualitative research, I have not generated any hypotheses.

My main question is, can I run other tests like t-tests, correlation, univariate, and Regression, to achieve outcomes like comparison between the IDV and effectiveness without hypotheses?

Do you mean compare the means to a wider group within my sample, or compare it with other studies?

Okay. Let's go back to your original intent: "I intend to explore the relationships between these variables (leadership effectiveness) and the five components derived from factor analysis . . ." You use the word RELATIONSHIPS. The best measure of relationship is correlation. That will allow you to measure the strength of the relationships, not cause and effect. There is no need to develop hypotheses for correlations; mainly due to these results will not be cause and effect. The five factors will result in measured strengths. I have attached a table I have used in the past. You also can look at the Multifactor Leadership Questionnaire (MLQ) for possible means on like items/factors and compare means to your results (one sample t-tests, or other possible tests - that will require consultation with a local statistician). Those tests would have hypotheses, e.g. There will be no significant differences between the MLQ Mean and your Mean, etc.

https://www.mindgarden.com/documents/MLQGermanPsychometric.pdf

Good luck with your endeavors. Dr. C

Thank you Mr. Collins for your time, patience, and the help you offered.

Let's approach this inferentially in steps like this:

1. Use the total effectiveness as the overall mean of all the variables utilized to evaluate or assess their leaders.

2. Conduct factor analysis.

3. Perform a correlation to identify these dimensions.

4. Execute regression using the mean of total variables as effectiveness (DV) and the five factors as IDV. What other inferential tests should I run? Regarding MLQ, my constructed questionnaire contains different items (variables) from MLQ. How can I use MLQ outcomes? Even if I use it, MLQ is typically conducted in Western countries, while my research is in the Middle East. It is not feasible to use or triangulate this kind of comparison due to cultural differences and other factors. I look forward to your academic advice.

Good plan. I shared the MLQ as another instrument for you to compare your results both qualitatively and quantitatively, as it is well researched (warts and all). I am sure there are cultural differences but some of the themes are likely similar. To compare individual items/questions (e.g. one sample t-tests) would require recoding as you use a 7-point scale and they use a 5-point. You are now well on your way to finalize your research. Again, Good Luck! 😊 Dr. C

Thank you kindly Mr Collins. Sorry again to trouble you. I need some declaration about using MLQ because I feel you are right to strengthen the outcomes by comparing my result with which achieved before with MLQ. Is it possible in one paragraph give me example of that. if I use ethical traits as a dimension/a factor that consists of integrity, fairness, and trustworthiness. what can I compare it to or how to compare it using MLQ (relevant factor or items from MLQ)?

Also, I read that I need to formulate hypotheses, as my research is a PhD thesis, even if it is an exploratory study, this formulation should be based on the previous theories. Is this right?

Your NULL hypotheses will be: There is no significant difference(s) in the relationships of five factors (IDV - list each separately) as measured by your questionnaire/instrument (DV). Like the MLQ, you can have four constructs (scores) for leadership effectiveness (competence, effective leadership, empowerment, and decision-making) or use a total score.

As this is your PhD thesis/dissertation you should be addressing these ideas with your mentor/advisor for approval. He/she is the best guide, as you will ultimately have to defend your thesis/dissertation with them. At least that is how it works here in the Unites States. I have been on over 160 doctoral committees and completed two myself (Ph.D. and Ed.D dissertations) and this is fairly consistent. Good Luck! Dr. Collins