Thursday, June 17, 2010

How can I use MAXDictio?

There is a lot to MAXDictio that we do not understand yet. For instance, it appears that you could use it to do some advanced coding. However, this will be for you to figure out and add to this tutorial later...

Word Frequencies (+ Autocoding)
********************************
One way MAXDictio can be used is to examine word frequencies. This can be extremely helpful if you are wanting to see what words are popping up the most in people's responses.

To use this feature...
  • Click on "MAXDictio" on the menu bar
  • Select "Word Frequency"
  • In the window that pops up, you can sort according to frequency if you would like
  • Just click on the header that you would like to sort by

Click on the image to view it larger. This is what the "Word Frequency" window will look like.


  • You can use this list to read through some of the responses if you wish.
    • Just right-click on the word
    • Select "Search results"
    • Another window will appear that contains all the instances of this word in your document.
You could also then autocode these responses if you wish (while in the "read through" screen mentioned above). Refer to the other tutorial for how to autocode.

Wednesday, May 5, 2010

Our Qualitative Coding Process

  1. Each coder should separately read through 100 - 200 responses and create what he or she believes to be the categories of responses that are revealed.
  2. Get together and compare/contrast categories. Decide on a set of categories that you will then both use to code.
  3. Code 200 - 300 responses using these (separately). You may also create new categories as you go along if you deem appropriate (e.g. new things are popping up that you know really don't fit into your already decided upon categories).
  4. Get together again and compare/contrast the frequency of responses each of you now has in each category and also the actual responses that are being placed in each category. Reconcile differences.
  5. Talk about any new categories that you feel are popping up that need to be new categories. Decide to add them or incorporate them into already existing categories.
  6. Code some more. Get back together. Repeat...

Wednesday, April 14, 2010

SPSS Text Analysis

Getting your Data Ready for Text Analysis

Most of our open response data comes from SPSS documents, and text analysis claims to be able to import your items directly from an SPSS data document; however, it has been my experience that this never works. You will need to then go into your SPSS data, select the variables that you want to export, and then export them to an Excel file.

Here is what the Excel file should look like (click to enlarge):


It needs an ID column, and each column after that would be another one of your open response items (with the responses following below each title).

Importing your Data into Text Analysis

Note: You will want to save your Excel file (and all of your further files from here on out) in one folder that you will NEVER move. The reason for this is that if you ever move one of the files (even move the folder that contains all of the files to a different place) it will break all of the connections that have been set up between Text Analysis and your Excel data file. This will cause you to lose much if not all of what you have done (even after you try to have the program rebuild the connections).

This tutorial was not continued because we switched over to MAXQDA.

Tuesday, April 6, 2010

MAXQDA Tutorials

 How do I begin?
**********************************************
After opening the program, you will need to import a text document for you to code. In order to do this, you will need to convert it from whatever form it is in into a simple text document (.txt). Most of the data we work from comes from SPSS, so my video tutorial shows the easiest way to do this from SPSS.

Importing a Document
**********************************************
I had a lot of trouble trying to import documents. For some reason, any kind of file (no matter if I followed their directions or not) failed to import. Thus, I recommend converting your file to a simple text document first (.txt). Then, click-hold and drag the document into the window labeled "Document System." If the import is successful, you will see the document appear in this window.

Click here if you can't view the video below.

Note: It views best on full screen mode and on the highest HD setting 1080p, which is available near the video player's right hand corner after you click play).



Creating Categories
**********************************************
Now, click-hold and drag the document (that you imported previously) from the "Document System" window over to the "Document Browser" window. If this is successful, the text should appear.

After you have read through enough responses to get a feel for what categories should be created, you can create your categories. To do this, right-click on the "Code System" inside the "Code System" window. Then, click "New Code" on the drop down menu that appears. You can now type in the name of the category and you may change the color of the tag. Click "OK" when you are finished here. You have now created your first category. To create others, just repeat the process.

Coding Responses into Categories
**********************************************
To code a response into one of your now created categories, use your cursor to highlight the exact text in a response (in the "Document Browser" window) that you believe fits into your category. Then, click-hold and drag your cursor to the category in your "Code System" window. If it was coded successfully you will see that the number to the right of the category now says "1" instead of "0".

**When selecting the text to categorize, it is generally best to select the entire response. This will help with duplication when you autocode later. A quick way to select an entire response is to click on the number to the left of the response.

Note: The video views best on full screen mode and on the highest HD setting.



MAXQDA - Video 2 - Creating Categories and Coding Responses
 

Creating Subcodes
**********************************************
  •  Right-click on the category where you would like to create a subcode
  • Select "New Code"
  • Type the name and select the color that you would like to be placed with this new subcode
  • Done!

Text Matching and AutoCoding
**********************************************
 Note:  If when you go to autocode you can't select the category that you want. You will need to get out of the autocode box, select the category on your main screen, and then start over. It should now pop up as available when you go to autocode.
  • Click on the search (magnifying glass) on the top menu bar
  • Create your search criteria
    • Be careful to use "And" and "Or" appropriately
  • It will then pop open a list of all of the responses that contain that text segment.
  • You can highlight whichever responses you want and then click on “autocode”.
  • Then, select the code that goes with this search.
  • Click "Autocode" (Note: Make sure that "paragraph" is selected).
  • It will code all that you selected into this code!
You will want to keep track of your autocodes in a memo on the code/category (see "Creating Memos").

Creating Memos
**********************************************
  • Right click on the code
  • Select "Code memo"
  • Write your memo and then close the pop-up box

How can I see what I  placed into a code?
*********************************************************
  • Right click on the document you are in (up in your "Document System" window)
  • Click on Activate
  • Then, right click on the code in your (in your "Code System" window)
  • Click on Activate
  • The coded segments should have appeared in your "Retrieved Segments" window
  • To get these to go away and to look at another just right click and deactivate the code. Then, you can activate another.

Eliminating Duplication
**********************************************
Note: When you code, code entire responses (sometimes called "paragraphs") instead of individual text segments within responses.

Before you export your frequencies, you will need to check for duplication inside your codes. This can be done on each individual code. Let's just start on our first code...
  • Right-click on the code/category
  • Click on "Overview of coded segments"
  • In the next screen that pops up you will see what has been coded into this category/code.
    • Click on "Begin" in order to sort your coded segments
    • Now you can see if there are any duplicates coded.
      • For instance there would be a 126 on one line and then a 126 on the next line.
  • You can delete a duplicate code by right-clicking on that line and clicking delete.
  • You can then repeat this process for each of your codes/categories

 Exporting Results/Frequencies
**********************************************
Make sure that every response has been coded at least once!
  • Click on "Codes" in the menu bar
  • Select "Frequency of Codes"
  • Click on "Export" at the top of the window that pops up.
  • Name it and Save it where you would like
  • It will ask, "Unicode text?"
    • Click "Yes"
  • It should open up in Excel. Now, just format the frequencies into table form.
  •  NOTE: When calculating the frequency percents do not use the total codes number that MAXQDA gives you. This is the number of coded segments, not the total number of responses in the data set. You will have to manually calculate the total number of responses. For instance, you could go into your document and see how many responses there were. Let's say there were 4,500, but you also know that you had blank responses in there. Those shouldn't be counted in your total, so you will need to subtract those. That means you might end up with a total of 4,200 after you get rid of the blank responses.

Welcome to the IAA Blog!


This blog is dedicated to providing tutorials, hints, documents, and other helpful information to all BYU IAA staff. 

Feel free to browse for what you need.

If you have any questions contact btmcdaniel2007 (at) gmail (dot) com or steve_wygant (at) byu (dot) edu.

Thank you.