Видео

Have you watched the earlier videos in this series? My guess is that you kept factors that have no sorts flagged. You should only keep those factors that have sorts identified with them. I should add that Q methodology rejects the idea of scaling so what you are doing is a Q methodology study, not a "Q-sorting Scale Development."

It's that you want to structure the Q-sample to ensure representativeness of all types of communication on your topic. My book doesn't come out until early 2025 but certainly Brown's 1980 text (PDF available FOR FREE from www.qmethod.org) has excellent coverage of this topic.

This video is a little old but the process isn't so different now. ruclips.net/video/77yho_PwdOM/видео.htmlsi=TLT5rbywQ9WDUpk2

@sue-zq9058 ohh thank you for answering Dr Sue! But my question still stands, I have the latest tool with EQ configuration and EQ websort. I created the sort materials, did the sorting and sent it but I don't know to where it goes. I chose netlify as server at the start.

@@PhDChaka the data is stored in Firebase, not Netlify. To export data as JSON: Select the node you wish to export. This also exports all children of the node [this just means all of the sorts within the database]. Click action menu and select Export JSON. Your browser will begin downloading a .json file for your database data.

@@PhDChaka Netlify is the server but if you followed the EQ configuration steps, your data is stored in Firebase. Sign into Firebase and look for your project. You should see the data there.

@@sue-zq9058 Hie Sue! Thank you for the follow up. I followed everything you did in EQ Configurator and even followed the Firebase manual, which i completed. However, when I copy the local host link under and paste it to my browser, its not opening. Its strange because before I saw your video, i had done the things you did using EQ Web Configurator (as in this video - ruclips.net/video/Whn7Bo2TSc8/видео.html) and when I copied the link and pasted it, it was working and I could see my statements, place them where I want but could not find where to retrieve them. Now I can't even see the start page of the survey! As it stands, I have my instructions entered in both EQ Websort and EQ Configurator. For the former, I can paste the link and it works. For the latter, the link I paste is not opening any page. So my question is, what is the difference between the two? And can me having instructions in EQ Websort tool affect the instructions I have fed into the EQ Configurator tool?? Which of the tools is more up to date? This is so confusing. For a very important methodology which people are trying to get into, there isn't much information to work with :(

The online database used should allow you to download a JSON file. This file can be read with software that reads simple text such as WordPad or Notepad [in Windows]. You will see each entry as a combination of survey question responses and the sort submitted. For the analysis in KADE or Ken-Q, only the sorting data is accessed. You will need to collect the survey data via the JSON file. Sometimes, if you used a multiple-choice format for a question, you will only see a numerical value for the response. This numerical value corresponds to your list of choices in the multiple-choice question.

@@sramlo Thank you for your quick and kind response:) Unfortunately, it did not works...TT I tried all of the database options (netlify, google sheet, firebase), but I cannot find the survey answers.. there are only data from the sorting process.

@@jinkyung412 Did you download the JSON file and open it using a text editor like Notepad?

@@sramlo yeah, I did! and I check the result of sorting but not for survey data..

@@jinkyung412 so, when you open the JSON file in Notepad, what do you see? Can you copy and paste the first entry here?

There are a couple of possibilities for this. I found when posting a link to my Q study via social media that participants had a similar error. Do you know what browser they are using? Does this happen with all participants or just some participants?

just some participants, there are safari or firefox, when it happened, i lost the participant😞@@sue-zq9058

Go to Config and then pre-sort options. That's where you can change the colors.

I agree!

Did you choose 7. Analyze after the rotation? You have to do that to create the .lis file.

Also, you need to flag the sorts on the factors

Thank you!

Hi Brittany... I'm not familiar with the output of VQMethod but if I had to guess from what you describe, I'd say that the format from that software / database download isn't quite the same as what KenQ expects for Type 1 data. I don't use the MS Excel data upload option so I'm afraid I'm no help as far as the intricacies of the data placement in MS Excel. You might want to contact Shahab Nazariadli about the format of the MS Excel file from VQMethod and its compatibility with analysis software.

@@sramlo So kind of you to reply, Susan! I've thought I've solved it a number of times in placing all the data directly into KenQ templates from the original Excel output, but haven't yet, so I may reach out to Shahab--he was so helpful when I first set up my study. Happy Thanksgiving!

Go to the Firebase database. You should see submissions listed. Hover your cursor over the identification number shown - it should turn blue and you should see a trash-can... click on the trash can to delete that entry.

By sample size, do you mean the number of statements in the Q sample (in Q, this is the sample size, not the number of participants in the P-set)? Z-scores are appropriate here although when reporting the factor arrays for the factors, I usually use the grid positions rather than z-scores. This is easier for especially non-Q researchers to interpret / understand.

@@sramlo I mean number of participants that it will interpret to Z-scores that sig or non-sig.

Try removing the return after the last statement in the .sta file. That's my best guess about the mismatch between the number of statements and the Q-sort design.

@@sue-zq9058 yes I did that, but the same error. I suspect it could be the format of the Q-sorts. Mine are listed by Q-sort codes (row) and values by statement numbers (column). I just couldn't fund an example of a .dat file, ifyou know of any. Thanks

@@fitsumweldegiorgis9404 every PQMethod download includes a set of files including the .dat file... I would look there for an easy test of your files.

Not that I know of but the good news is that the HTMLQ JSON file can be uploaded directly into KenQ for analysis. So they aren't a single package but they are interconnected in a way.

Yes... it is only the data "intake" that is different but everything else remains the same.

@@sue-zq9058 Thank you very much from south korea! 🙏

Thank you for your kind words!

Hi Zain, may I ask how you are getting yourself trained? I am new to Q-methodology and trying to learn. Sue-Z Q, your work is amazing and so very helpful! Thank you

If I had to guess I would say it has something to do with the database setup... it just isn't accepting what is being submitted.

Dear Stijn, Currently I'm facing the same issue. Did you managed to fix the "data submitting error"? Kind regards, Thom

@@thombongaerts Helpful video - thank you. I'm also having the same issue - Nothing happending when I click submit

Probably not worth your time to do this within PQMethod. But KenQ is perfect for uploading a MS Excel file and then doing analyses... see my videos about using KenQ!

@@sramlo KenQ is sort of pq method?

@@비비비1 Yes in functionality they are the same but KenQ is not a DOS based program but one that runs on any internet browser. KenQ is more modern.

@@sue-zq9058Thanks, as far as i know, pq method is better than Qunal. If you say so, KenQ is most uptodated; it would be better than any others. Thanks! again!

Sure... sramlo@gmail.com is probably the best email (with the new normal my work email is usually blowing up with emails).

Not in PQMethod but because I make everything via MS Excel (accessible at uakron.academia.edu/SusanRamlo) it is easy to put the factor arrays into the grid I created (in Excel)... and I typically do something like that in publications... not for all of them but usually for a key viewpoint.

Yes! I do it all of the time.

That's a great idea. Do you do that within the online classroom management system your institution uses?

Well, a z-score represents the number of standard deviations from the mean. You can use it for nonparametric distributions (e.g. a left skewed distribution). Looking at your example sort in your post, I would say the distribution is fairly close to a normal distribution. You need to have some kind of reference point for discussing your data. Frankly, in my most current study I had multiple participants not follow the grid I provided for sorting -- but when I discussed my data in my presentation a few weeks ago I referred to the factor arrays for each factor based on the grid I presented. I have not used free-sort as a technique mostly because I want to "force" my participants to make what they might perceive as difficult decisions. For the recent study I mentioned, those who deviated described their thought process as difficult and certainly had thought about their placements outside of the grid very carefully. I hope that helps.

See reply above in other thread...

Hi Muhammad and Susan, Thank you for your discussion on free sorting. This really clarified some questions I had. I ran a practice workshop in April and several participants were not happy that they could not stack most of the statements in the -5 column. I explained that limiting the -5 and +5 columns to two statements forces them to make a ranking choice. I might shuffle the columns a bit for the next sort to allow for more stacking options.

Hi Muhammad, If you look at the PQManual that was included with your PQMethod download (it is called PQManual.htm) you will find a description about how the z-scores and factor arrays are calculated. The grid values are based on a rank ordering of the z-scores (in short). There are a variety of tables in the LIS file including tables with both z-scores and factor array values... I have found that using the grid positions are much easier for readers to comprehend -- and it is easier to relate those values to the sorts provided conceptually. Does that help?

Hi Susan E Ramlo: Although i defined grid 2 3 5 7 10 7 5 3 2 , but as i used free choice sorting i put zero where or left the column empty where respondents didnt not place any statement for example: i,e Sorting by one of the respondent, is shown: the rating scale ranges from -4 to 4, -4 column do not contain any statement -4 -3 -2 -1 0 1 2 3 4 41 8 34 36 27 15 25 1 14 35 42 28 16 2 9 31 32 30 17 3 20 37 33 44 4 22 21 38 10 18 5 26 43 19 6 24 7 39 23 40 The confusion is that i have used free choice sorting method (giving the respondent freedom to rank any number of statement in any column), so it means data do not follow normal distribution (not normally distributed). what i have read from many sources that "z-scores" are only calculated for data that follows normal distribution and cannot be used for non-normal distribution. The PQ-Method manual states that 1) The exact computational procedure consists in first z-standardizing every sort. 2) Then applying different weights for every sort depending on the sort's factor loading, and computing the weighted average. 3) Finally, every factor score is z-standardized again, i.e. every factor score has the same mean (0) and standard deviation (1), How it is possible for Pq-method to calculate z scores or standardizing z score when the data do not follow standard normal distribution.? Please help and guide, its very important for me

But you had to enter a distribution for PQMethod at the beginning.... so I would reference that... how else will you interpret the findings holistically if it isn't based on a distribution of some kind?

Although, PQ method requires to define a normally distributed grid, but the data in the grid is not normally distributed surely, Hi Susan E Ramlo: Although i defined grid 2 3 5 7 10 7 5 3 2 , but as i used free choice sorting i put zero where or left the column empty where respondents didnt not place any statement for example: i,e Sorting by one of the respondent, is shown: the rating scale ranges from -4 to 4, -4 column do not contain any statement -4 -3 -2 -1 0 1 2 3 4 41 8 34 36 27 15 25 1 14 35 42 28 16 2 9 31 32 30 17 3 20 37 33 44 4 22 21 38 10 18 5 26 43 19 6 24 7 39 23 40 The confusion is that i have used free choice sorting method (giving the respondent freedom to rank any number of statement in any column), so it means data do not follow normal distribution (not normally distributed). what i have read from many sources that "z-scores" are only calculated for data that follows normal distribution and cannot be used for non-normal distribution.' As i have used free choice sorting method and where respondent has the freedom to put any statement under any column. and as advised by the PQ-method users i just put a '0' in the column in which user didn't placed statement and hence distribution is not normal, . Can you please advise on that

See my response on your same comment below... I hope it helps!

Awesome... glad they are helpful! Did you chose an extraction first? Either Centroid or PCA (I'm a centroid person -- it is what Stephenson preferred for reasons that fit within a quantum mechanical philosophical framework...).

No it is not required. It is not unusual for my participants to not conform to the grid. You need only enter their sorts as they provided them. PQMethod will tell you that the mean does not equal 0 but that is only a notification -- not something to worry about. It will not negatively impact your analyses.

Thanks for this ans. It cleared one of my big confusion. One more thing, i have chosen free choice(unforced) sorting what if any participant didn't loaded/ Placed any statement on lets say -4 and 3.. Do i need to discard that complete response of that particular participant? Or just fill that column with a value of zero and move along to next column..? will it negatively impact my analysis? Secondly if i have defined my column "-3" to take 4 statements in PQ-method and someone has ranked 6 statements in that column i.e "-3", is it fine to put these 6 statements number in that column and move along? or do i need to discard the complete response of that participant? Remember i am using free choice sorting the ques might appear very basic but i am a new-bee with Q methodology, Please answer this query. Really need this

It's ok... enter the sort as they participant provided... if there is no statement in a box (position) then leave it blank... the only real requirement is that every number needs to be entered only once... if they are missing say 1 or 2 numbers of statements sometimes I just place those in the 0 column and accept their sort that way.

That's fine... just enter the sort as the participant provided it. A deviation from the grid will not negatively impact your analyses or results.

Muhammad, you will still need to define a grid when you go through the steps of the PQMethod software... your entries will just not necessarily conform to that grid which is just fine. Just enter the sorts as they are provided by your participants.