Webinar - Scale Development and Validation: A thorough guide on how to develop and validate a scale
HTML-код
- Опубликовано: 21 авг 2024
- The session focused on Scale Development and Validation. The session focused on
Key Steps in Scale Development
Exploratory Factor Analysis
Confirmatory Factor Analysis
Scale Reliability
Construct Validity
#ScaleDevelopment #EFA #ExploratoryFactorAnalysis #ConfirmatoryFactorAnalysis #Reliability #Validity #ConstructValidity
You can't imagine the scale of help you've provided to me Dr. All love, respect and regards from Egypt to you 🇵🇰🇪🇬🙏
I am very happy to know that the video was helpful.
VERY useful and informative. Thank you very much!
Thanks for your interest. I am glad you liked it.
Thank you so much, sir. Your video has cleared almost all my doubts. God bless you.
Thanks. I am glad you liked it.
Hello Dr Latif. I loved your presentation, it is very clear and helped me understand scale validation. Thank you very much!
Thanks Ana. I am glad that it helped. Do circulate the channel in your research circle. Someone may benefit.
Sir, you have made my day...Excellent explanation! very easy and understandable. May Allah Pak Bless you sir!
How to contact you sir for any guidance related to thesis and research work? Is it possible?
Thanks A. Ahad. I am glad it was helpful.
You can email me on kh.fawad83@gmail.com
@@researchwithfawad Thanks a lot! I will send you email Inshallah
Thank you.. this video and concept means alot to me.. thank you alot
Most welcome 😊
Thank you so much Dr.
Pleasure. I am glad you liked it.
Thats really very easy way to explain
Thanks. I am glad you liked it.
Thank you for replying to my query earlier that I can add statements to already tested scales of previous researchers but then recommend validating it for content and face validity, and running EFA on the scale.
Sir, can you please elaborate on how to go about or share videos/resources . Will be immensely helpful !
Hi. For EFA watch this ruclips.net/video/ust9X0CKKTk/видео.html
Reason i am suggesting efa is because you are incorporating new items and this may change the factor structure.
For face Validity get your scale reviewed by a senior prof or any ogher subject expert who has researched in the area
Its just amazing.....
Thanks. I am glad you liked it.
Wonderful
Thank you. I am glad you liked it.
Thank you for your such a comprehensive video. I was wondering how must we treat multiple missing scores?
Thanks. I am glad you liked it. For missing data watch
ruclips.net/video/dZdIiEsgHWE/видео.html
Fantastic - thank you!!
Thanks. I am glad you liked it
What a wonderful video sir. Thank you so much. Love and respect from India.
Sir, I have one request. Can you please provide this data to practice? and also the research papers you showed here?
Thanks for watching. Sample data is available on researchwithfawad.com
@@researchwithfawad Thank you so much sir.
Thank you so much sir
Most welcome.
Excellent! You illuminated a lot of areas, and did it without miring us in formulae and in a relatively short amount of time. Nice job. Now I have a question, some psychometrists correlate their scale after EFA with known correlates of the main construct and then remove items that don't correlate sufficiently with known correlates. For example, Boredom Proneness correlates with Aggression. Boredom Proneness Scale items that don't correlate well with Aggression are removed. What kind of validity would this be called? Concurrent validity?
Thanks. I am glad you like the video.
For for concurrent validity, the measure is correlated with the well established assessment/measure. Here in your example Boredom and Aggression are different concepts. This can be referred to as nomological validity, the is focused on testing the relationship or predicting a construct of interest.
Thank you very much! I assume you mean the concurrent validity is associating the scale with another, established scale that measures the same construct -- right? Nomological validity means testing the scale to see if it associates with constructs that are known to associate with the construct you are testing. Correct?
Yes, please.
Hi Fawad ! Can you please provide the AVE Calculator and other calculators that you used for the analysis? Thanks in advance :)
Thanks for your interest. Please visit researchwithfawad.com for download. File is available in downloads section.
dear brother really impressive information. As i am developing a construct for my variable. i need your assistance. will you please help me?
Thanks for watching. How can I help you?
Thanks for the wonderful video on scale development. I have read your papers as well. If possible, could you prepare video on before run EFA because there are so many variations of techniques such as interview, FGD and herarchical clustering.
Great suggestion! I will work on it. I guess i can make a video on Steps before running EFA in Scale Development.
@@researchwithfawad Indeed that would be awesome!
Thank you for this presentation. I was wondering how you can determine the order of the items within the scale. Is it better to group them based on subscales or present them randomly? Does item order need to be fixed before analyses? And, once the scale is validated, can item order be changed (e.g., if items were randomised during data collection). Thank you!
Hi. Thanks for watching the video. I normally keep items of a single scale together. I do not randomize the items. Although, randomization may help.
Thank you so much for such a wonderful presentation. Could you please share link of our publication.
Please email me on kh.fawad83@gmail.com.
I will share the paper
Thank you so much for your comprehensive video. I just have a query in EFA. What if some items are loading negatively in the rotated component method?
Please. This is because the item is negatively related to the factor.
Dear Dr. Latif, Thanks for your very clear presentation. Just wondering whats the name of the software you used in this demonstration?. Would you please share the link to your publication you used in the presentation? thank you very much
Thanks for the interest. The software is Jamovi. It is free to use. Drop an email on kh.fawad83@gmail.com. I will send the paper. Thanks
Great Sir very well explained... Would you provide me ppt of this Webinar if possible please?
Let me search, will email if i find it, i hope it was not deleted during that virus attack
Thank you for the presentation, Dr. Latif. It has been extremely helpful. I was wondering if we are supposed to conduct CFA on a different dataset than the one used for EFA? In your example, I believe you used the same dataset if I did not miss any point. Also, what is the difference between Principal Component Analysis and Principal Axis Factoring? Are there specific situations where we are supposed to use a specific one? Thank you.
Thanks. I am glad you liked it. For CFA, although preferred is to have a new dataset. However, researchers have used the same dataset. I too have used the same in most of my studies.
As for PCA and PAF, both are dimension reduction techniques, one is focused on
Whether the items reflect or cause the factor. Read more.
stats.stackexchange.com/questions/280746/what-is-the-difference-between-pca-and-paf-method-in-factor-analysis
Thank you for the great presentation. I am conducting a scale validation study using Jamovi inspired by your presentation. For Exploratory Factor Analysis, the results were better if I run PCA than ran EFA in Jamovi. However, one of my coauthors pointed out that we should not use PCA for EFA since they are different. Could you please provide a valid justification for why we can use PCA for EFA? Also, the variance explained by my data is only 54% after selecting the appropriate number of factors, is this okay since most references mentioned that this value should be at least 60%? Thank you so much.
Thanks. I am glad you liked it. if through scale development, your focus is on reducing the number of variables while retaining the most amount of information in the data, PCA is useful.
Thank you very much.
Sir if we take different constructs from different authors and also in some constructs combine two or more authors and make a scale can we use Efa in this regard ?
Yes, better to use EFA
Can you provide the spreadsheet you used to calculate composite reliability and AVE?
Thanks for your interest. You can download them here
researchwithfawad.com/index.php/download-research-resources/
Assalam o Alikum sir hope for your good health.. I wanted to ask.. Removing items during factor loading means discarding them from scale???
WAS. Yes, you have to remove them from further analysis and the final scale.
Thanks sir.I need this lecture in PDF form
Please visit the link to download
researchwithfawad.com/index.php/understanding-the-scale-development-process/
Hi Dr. Can you share a link to the paper you were referring to for efa and cfa
Please can you mention the title from the video.
should we first do validity or reliability? everyone has different opinion. some say reliability should be first , while others say validity comes first. Since from social science background, different opinion confuses me. please clarify the order of analysis sir. Is there any reference for order of analysis?
Factor Loadings, Reliability, and then Validity.
@@researchwithfawad thank you so much sir.
Hello Sir, can you make one video on the difference between scale, variable, and construct, and in case it is already there can you please share the link. Thanks once again
Hopefully soon.
Greetings sir
If we take different statements from different studies and form a scale then it will be adapted scale? Can we apply efa in this regard.
Yes it is an adapted scale. Better to check it with EFA
great video sir... Thank u so much... May i ask if u have a video that uses rasch analysis in scale validation?
Not yet. I will work on that and will upload a video soon.
@@researchwithfawad thank you sir... It will be a very big help for us..
Sir could you make a video about scale modification from A -z
Thanks for the suggestion. Hopefully soon.
Thank you sir! Sir, can I adapt the scale of previous researchers, say done in the logistics industry to use in other service sector industry? Can i add more relevant items and what should be the next process to make it relevant?
Yes you can add statements but then i recommend validating it for content and face validity, and running EFA on the scale.
@@researchwithfawad Thank you so much sir for the help and prompt reply ! I really liked your presentation, clarity and explanation ! Sir, may I contact you for any guidance ?
After principal component analysis, the component loading matrix show items from different construct coming in a single factor and the number of factors are more than constructs. The items you have generated for each construct very well come under a single factor. Can you please suggest a remedial action for this as well?
If items are not loading onto their respective factor, those items shall be removed one by one, or if the items are not loading acceptable loading they shall be removed too. But if you are using an already validated scale, do not go for EFA, go for CFA
@@researchwithfawad I'm not using an already validated scale so I've to for EFA. Does the sample size has an affect on this? Because this was observed during pilot test.
Sir, can we use a multidimensional scale into a unidimensional to avoid the model complexities ?
Yes but is there a theoretical justification or analytical one
@@researchwithfawad Analytical one. When I tried to see the impact of this variable treating it as multi-dimensional then all of the results were insignificant but when I treated it as unidimensional, the results were quite significant.
Gud day sir! I watched ur video a lot of times.. it is very helpful sir to my study.. i am just curious sir if I still need to validate the questions for my Focus Group Discussion in the developmental phase of my study.?.
Hi. Thanks. If you are developing a questionnaire that is based on interviews/focus group and literature, the questionnaire shall be validated.. Even if you are using an already developed scale, Reliability and Validity must be tested.
@@researchwithfawad ok sir... Thanks.. what i mean sir are the questions for focus group discussions/interviews that will be asked to potential respondents and will be used to generate the items. .i still need to validate those?
Sir ! Do you have good references for scale development research papers? Easy to comprehend for a beginner.
Thanks for your comment. This paper may be helpful
Title: Measuring corporate social responsibility: A critical review of
survey instruments
Scale development i a complex process, reading more and more will help. Try to read papers on scale development.
How to remedy low KMO value? During a pilot test of my questionnaire, I observed the KMO value is 0.28 which is below the threshold value 0.5.
KMO is a measure of Sampling adequacy, your sample is low. We do not normally use EFA in Pilot. If you are using previously validated scales you do not need to conduct EFA.
@@researchwithfawad Thank you , Do you have a video of testing hypothesis? For instance, testing the hypothesis that psychological factors are a significant dimension of consumer behaviour? or something similar which studies causal effect?
For doing first pilot study of rating scale how many samples we should take
For pilot a sample of 50 is good enough. But dont put it through EFA, just check it for reliability.
WHICH QUESTIONNAIRE TO BE ADDED IN THESIS WITH FULL LIST OF QUESTIONS OR FINAL LIST OF QUESTIONS AFTER EXPERT VIEW
The final questionnaire used to collect the data.
@@researchwithfawad HOW THE ITEMS ARE REDUCED BY WHICH TEST AND WHAT IS VALUE BELOW WHICH ITEMS ARE DELETED .
CAN I CALL U .PLEASE HELP ME SIR .
Please watch this
ruclips.net/video/ust9X0CKKTk/видео.html
Sir, does development of any scale , test or inventory depends on the variable that we choose .if i want to make any one of three(test/scale/inventory) then it will depend on variable or my choice ?🤔
Can you please clarify your question, please
@@researchwithfawad suppose i choose a variable. How do I know whether this variable will scale or test or inventory? . Beacuse test ,scale and inventory are different from one to another.
Or
Can we make all (i.e.test / scale/ inventory ) on a variable?
You can possibly look at the original source and see if the development was for a scale or something else
Warmest greetings, Sir! Thank you for this great help. Sir, should I still go for EFA even if I already modified the adapted instruments from the existing study?
Pleasure. I am glad you liked it l. No you don't have to perform EFA for an established scale that is already validated.
@@researchwithfawad Even if I added some items to it? Then, I should proceed to CFA already? Thank you for your quick response.
How did you add the items. Did you combine two scales into one. Even then, no need to do EFA
@@researchwithfawad I just added items. Is it okay to send my initial factor analysis for you to have a closer look? Actually, after I have finished your video, I also tested each dimension separately to get a better result.
Thanks Dr
How can I make the key score that help me to measure the level of X construct such( high , moderate, low level) among group of students?
You can yake mean ans then take percentiles or define what you can refer to high, and low, while moderate could be your mean.
@@researchwithfawad
Could you please provide any support or reference?
pls share the link of this paper, pls
Please email the title on askRWF@outlook.com
Can i get the article to study
Please mail me on kh.fawad83@gmail.com. I will share the article.
Sir I like your presentation
If you can read my comments please help me on my Dissertation..
If you have any questions, you can email me on kh.fawad83@gmail.com
What’s ur email?
kh.fawad83@gmail.com