HelpForum

Forum (Archive)

This forum is kept largely for historic reasons and for our latest changes announcements. (It was focused around the older EPPI Reviewer version 4.)

There are many informative posts and answers to common questions, but you may find our videos and other resources more informative if you are an EPPI Reviewer WEB user.

Click here to search the forum. If you do have questions or require support, please email eppisupport@ucl.ac.uk.

<< Back to main Help page

HomeHomeUsing EPPI-Revi...Using EPPI-Revi...New Feature Req...New Feature Req...Coding Results for synthesis of qualitative studies Coding Results for synthesis of qualitative studies
Previous
 
Next
New Post
02/06/2011 09:08
 

Hello,

We're coding the Results sections of articles in the 'Text' function of Eppi-R4 for the purpose of a synthesis of qualitaitve studies. The following features would be useful, but we haven't been able to figure out how to do them -

1. Upload files other than pdfs (e.g. Word or Html docs)

2. Keep the formatting of pdfs, which otherwise becomes a jumble of table contents, journal details, page numbers etc.

2. Delete irrelevand text (e.g. page numbers etc) or add notes;

3. Merge codes when one recognises that a previous distinction was not a good idea; 

4. Get an overview of what text has and has not been coded throughout the document to make sure nothing's been missed.

Are any of these possible?


Many thanks,

Tim

 
New Post
02/06/2011 12:47
 

Hi Tim,

thanks for sharing, we always look forward to this kind of feedback.

1. Upload files other than pdfs (e.g. Word or Html docs):
I'm puzzled about this one. The system already allows to upload the most common file-types. I must add that it relies on the file extension to understand what type of file is being uploaded and wether to accept it or not. This means that Mac users may need to manually add the file extension to their files before uploading them. For example: to upload a word (2003) file, one should make sure that the filename ends with ".doc", the ".docx" extension should be used for Word 2007 files, the ".htm" or ".html" extensions are both valid for HTML documents and so on.
If you'll stumble on a legitimate file type that you would like to upload but is refused by the system, please let us know, we'll try to start supporting it at the next update.
In case EPPI-Reviewer 4 is actually rejecting all your non-pdf files for other reasons, please let us know as well: we haven't had any complains about this so we'll have to investigate what doesn't work in your particular case.

2.a Keep the formatting of pdfs, which otherwise becomes a jumble of table contents, journal details, page numbers etc.
and
2.b Delete irrelevand text (e.g. page numbers etc) or add notes;
It is not possible to "extract" the text and keep all formatting, but yes, we are aware that the current solution is far from perfect. Our plan is to extend the native support of PDF files (only PDFs!) so that it will be possible to add codes to text directly inside the PDF document (instead of using the "text" tab). It is now possible to create PDFs from any printable document and in any platform (being Windows, Mac or Linux), and currently 98% of the uploaded documents are PDFs, so we don't plan to extend this feature to other formats: it does not seem a good use of our time. I don't know when this feature will be ready: we have written and/or validated all the necessary building blocks, but it is not something we can add gradually, one little piece at a time; we need to be able to write it all and test it throughly, so it requires more time and effort than other features.
As for editing the "extracted" text: we have discussed this option a lot and decided we prefer not to allow this. The extracted text is supposed to be a static "copy" of the uploaded document, if we allow users to edit it, we will have to solve two problems: first, we will need to handle text changes of already coded text, this can be more complex than it seems. Second, we will need to provide undo functionalities, or at least a "revert to original" function; one thing is sure, people will make mistakes and then ask us to pull them out of trouble. Combine these two requirements and you'll see why we have decided not to allow editing: for example, what are we supposed to do when restoring an already coded text?
Add Notes: there are two distinct ways to add annotations, one (preferred) is to add some text to the code itself through the "info" button, this comes with the limitation that the "info" text will apply to the "Item-Code" association, and not to the "FullText-Code" match. The other option is to annotate the PDF file itself: this type of annotation can be added to any part of your PDF documents but it will not appear on reports or the "text" tab, and is not associated with codes.

3. Merge codes when one recognises that a previous distinction was not a good idea;
I can see why this might be very useful in some special cases. We'll think about it, if we'll find a way to add this functionality without having to change the rest of the system, then it's likely that we will add it. Again, I can't promise neither if nor when, sorry!

4. Get an overview of what text has and has not been coded throughout the document to make sure nothing's been missed.
This is definetely on our plans, we have very ambitious ideas about this and that's probably why it's taking so long for us to have it ready. I expect this type of feature to appear after we will have finished the "native PDF coding": it makes sense for us to write something that will apply to PDFs directly - this is how we expect most people will code the full-texts in the future.

I'm sorry I can't give you more positive answers, please do continue to share your suggestions, though. We try to make sure that EPPI-Reviewer keeps evolving, and the drive from our users is the most important for us.

Best wishes,
Sergio

 

 
New Post
08/06/2011 08:59
 

Hi Sergio,

Thanks for your prompt response. I look forward to the new features as and when they come online. Merging codes in particular is quite essential in iterative phases of thematic analysis. Without the facility, we will have to copy and paste into a Word/Excel document and do second and future phases the old fashioned way. Can we run a report to see all text coded with a particluar code across articles or  will we need to do this one article at a time?

Re comparisons of codes run under 'Collaborate' - could you please explain what the codes listed under each reviewer in the report mean? Do they indicate only in general terms which codes have been used by that reviewer and not the other for example? On a line-by-line basis, it seems the report is missing some disagreements. 

Also, when reviewing disagreement in coding between two collaborators to decide which coding to accept, is it possible to view the text/codes of each reviewer side by side or do we have to have to switch between two Eppi-R screens (i.e. logged in separately as each person)?

Thanks,


Tim

 
New Post
09/06/2011 11:39
 

Hello Tim,

You can run a report to show all text coded with a particular code. If you are in the Document details window (where you apply a section of text to a code) and right click one of your codes you will see that one of the options are 'Report: all text coded with this code'. Selecting this will display all of the text coded with that particular code broken down by item. You will find this in the user manual under 'Inductive coding'.

If you set up a comparison in the Collaborate tab and run a 'quick report' what you are seeing are the codes that each reviewer selected (for the items and codes that make up your comparison). The user specifies what items are in the comparison and what codes or codesets that are compared so you should be seeing everything that you specify. You will find details on this in the user manual under 'Comparison and reconciling of multiple coded items'.

When you are reconciling differences between two (or more) coders you can turn on 'Live comparisons'. This will show what codes each person has selected for a particular question. This is turned on in the 'Coding record' tab in the 'Document details' window. You will find this in the user manual under 'Dealing with disagreements'.

Best regards,

Jeff

 
New Post
14/06/2011 05:16
 

Thanks for your response, Jeff. However, I'm afraid I'm still unclear and can't find the info I'm looking for in the manual.

From your first response, can I assume that it isn't possible to run a report of a given code across documents, rather only within each document one at a time?

Can you please explain what a quick report for collaborators actually produces? It can't be all the codes used by that collaborator because reports I've run for myself and a colleague show only 4 or 5 codes when we have used 20+. Is it the codes where there has been some disagreement? If so, how does the system identify these? It can't be a line as looking at our codings side by side, it seems there are more discrepancies than are listed? And is there a way for comparing our codings line by line other than logging into Eppi twice on the same computer and displaying side by side? The live comparisons I've run seem to present the same info as the quick reports, just displayed in columns under the text view.

Sorry if I've missed something obvious.

Many thanks,


Tim

 
New Post
14/06/2011 16:04
 

Hello Tim,

I must not be explaining this very well.

I just had a look at your review and I think I understand what you are asking now. I assume we are talking about your codeset called 'Coding for themes'.

I can see the codeset is set up for multiple data entry and there are two of you coding the same items using this codeset.

As this codeset has multiple levels of hierarchy some ways of running a comparison may be better than others.

Normally, in a multiple level codeset we would make our comparisons on an item by item basis.

To run a comparison at the item level you would want to be in the Document details window (where you carry out your coding). If you then click on the 'Coding record' tab you will see a record of all of the coding that has taken place on that item. For the item I am looking at I can see that two of you have both applied the 'Coding for themes' codes to the item. If you click in the checkboxes for the names and codeset your want to compare and then click on 'Run comparison' (not 'Live comparison') you will get a comparison, using different colours for each coder, of how each of you have coded the item. This comparison will look through all levels of the codeset.

If you are looking for something similar but across a number of items you would normally want to generate a report (in the reports tab). But for an items coding to be visible to these reporting functions the coding must be agreed and 'completed'. As you are still double coding the items and have not come to agreement and completed the items, they will not be visible to this function.

In this case you would want to use the 'Quick report' in the collaborate tab. The issue with this is that the quick report only looks down one level of hierarchy at a time. As your coding tool has a number of levels in it you would need to run a number of quick reports for each level/question. The behaviour of the quick report is similar to what you see with live comparisons where you only see the comparison at the level of the codeset you select.

In summary, with a multi-level codeset, to run a comparison before the items coding have been agreed and 'completed', you need to run your comparison on an item by item basis. If you want to run a comparison across a number of documents you would need to run a number of quick reports for each level of hierarchy in the codeset.

Best regards,

Jeff

 

 
New Post
21/02/2012 15:27
 

 Hi Jeff,

a quick query about 'inductive coding'. I would like to see the text that has been assigned to a particular code (code set type: multiple coding; coding not 'complete' yet), but cannot view it. When I repeat the steps outlined in the manual (p73, lines 11-12), I can see the applicable text highlighted in yellow. But, when I select the 'Report: all text coded with this code', nothing comes up. Could this be because the coding for this item is not complete yet? When I try doing the 'Coding record' tab and tick the code set - click on view, I see all coding done with the 'checkboxes' and 'info buttons', but not the 'inductive coding'.

thanks,

jan

 
New Post
21/02/2012 15:47
 

Hello Jan,

The coding needs to be marked as 'complete' for it to show up in the report. If your codeset is set for multiple data entry you will need to manually complete the item before the high-lighted text will display in a report.

If the item is not complete you can still select the option 'show text coded with this code' to see the selected text but it will only be for that particular item.

Best regards,

Jeff

 
Previous
 
Next
HomeHomeUsing EPPI-Revi...Using EPPI-Revi...New Feature Req...New Feature Req...Coding Results for synthesis of qualitative studies Coding Results for synthesis of qualitative studies


Copyright 2021 by EPPI-Centre :: Privacy Statement :: Terms Of Use :: Site Map :: Login
Home::Help::EPPI-Mapper::RIS Export::About::Account Manager