Importing ongoing survey questionnaires / filtering by foils

07.04.2016, 18:29

I have a question regarding importing ongoing survey questionnaires into MAXQDA. Specifically, what is the best way to filter by foils to closed ended questions?

I have an ongoing study with 75 participants. Every day, each participant sends in responses to a set of structured questions, open and closed (we are tracking how a particular thing fits into everyday life over a period of time). They answer the same questions every day for six weeks.

We are using a qual/quant mobile tool to capture the data, exporting it as an excel document, prepping the document to MAXQDA's specifications, and importing it into excel.

In MAXQDA, each participant is its own document (the title of the document is the participant's ID). All of the documents/participants are in one document group, "Cycle 1" (as we will be running six more cycles, each lasting six weeks). Scrolling down from the top, each participant's document includes the following:

-Their application to the study (mix of open and closed ended questions)
-Two submissions for a one-off survey (mix of open and closed ended questions)
-Two submissions for a different one-off survey (mix of open and closed ended questions)
-Up to 37 daily submissions (mix of open and closed ended questions)

That's the context.

The question I have is regarding my ability to filter the data by responses to closed-ended questions. I am having trouble doing this and I suspect we may not be importing the data correctly.

For example, one of the closed-ended questions we ask in the daily study is "was this experience positive or negative?" I would like to be able to filter the submissions by the answer to this question (e.g., pull up all 'negative' experiences to compare them). Currently, we've been importing this question as both a code and as a variable, with the hopes that we could both filter responses by variable and easily tally how many experiences were negative versus positive. When I pull up the data editor in the Variables section, only the very first submission is accounted for - I think because it is imported as a document variable, and so considers the document as a whole and only reads the first instance of the question, ignoring the rest. Unlike a variable like gender, which is presumably fixed, this changes from day to day. What can we do? This problem also arises with respect to a number of other closed-ended questions, including:

-Did anything related to this topic come up today? (Foils: Yes! / Nope)
-How long have you been engaged in this topic? (Foils: Less than a year, 1 to 3 years, 4 to 10 years, etc.)
-How has your relationship with this changed over time? (Foils: It has become stronger over time, It has become weaker over time, It hasn't really changed over time, etc.)

Perhaps they should not be treated as variables but as codes. If so, what is the best way to accomplish our goal of filtering by foils to closed ended questions?

Thanks in advance for any insight!

Version: MAXQDA 12
System: Windows 10
Posts: 1
Joined: 07.04.2016, 01:57

Re: Importing ongoing survey questionnaires / filtering by f

14.04.2016, 14:33

Hi, Many thanks for your post, which I have answered via our online support team. I post the answer here for others who have similar questions. Best regards, Stefan

Autocode daily submissions with object of fandom
As I see in the MAXQDA project, each submission starts with a submission date in the 3rd paragraph, then 7 paragraphs will follow. You can use this structure by following these steps:

1. Perform a lexical search for “/2016”, because each submission contains this phrase.
2. Autocode the result with a NEW code, let’s say “Autocode: Submission”
3. In the Autocode dialog choose minus 2 and plus 7 paragraphs

MAXQDA will autocode each submission now. This will be helpful for answering question 1 and 2. In a second step you can use MAXDictio to count the different objects of fandom:

1. Open “MAXDictio > Dictionary” and create one category for each object of fandom, with just a single search item, the object fandom, e.g. category “Harry Potter”, search item “Harry Potter”. (You can import a dictionary from an Excel file as described in the help ... naries.htm.)
2. Open “MAXDictio > MAXDictio Coder”, start it on level 0 and MAXDictio will count the number of search items per category per document.
3. Click on “Autocode with NEW code” and choose “+/- 0 Paragraphs”

MAXQDA will now autocode all search results for an object of fandom with the name of the dictionary category, thus the name of the object of fandom. If you use “Analysis > Complex Coding Query” with function “Overlapping” to retrieve all submissions for a special OoF:

1. Start “Analysis > Complex Coding Query”, choose function “Overlapping” and drag the codes “Autocode: Submission” and ONE OoF code, e.g. “Harry Potter”.
2. Click on the icon “Autocode with NEW code” in the “Retrieved Segments” window and code the results with a code called “Submission – Harry Potter”
3. Return steps 1 and 2 for all OoF

The number of coded segments serves as the statistic you are looking for. You can now use different functions of MAXQDA to compare the results for different people, e.g. the “Mixed Methods > Crosstab” tool.

Filter responses by answers to closed-ended questions.

As you have already stated, it is not possible to use the document variables in your case, because there are different submissions of one person in a single document. But you can use the previously coded segments of submission for the task, here demonstrated for “was this experience positive or negative”:

1. Reset the Retrieval functions to Simple Coding Search (there is an icon in the main toolbar)
2. Activate all documents and the code “Positive or negative” to retrieve all coded segments to this question.
3. Perform a lexical search for “Positive” only in the “Retrieved Segments” window
4. Autocode the result with a NEW code “Autocode: Positive”
5. Deactivate all codes and activate the newly created code.
6. Use again “Analysis > Complex Coding Query” with “Overlapping” to retrieve all “positive” “Submissions”
7. Autocode the retrieved segments with a NEW code “Submissions – Positive”

General hints

Usually it is a good idea to have several data for a person in a single document, because lots of analytical functions in MAXQDA work on documents level. At first I thought it would be a good idea to split up your documents into one document per submission and cover all documents of a participant in a single document group, but after I saw your project, it seems to be a better idea to keep it as you have designed it. Please note that the project may be very large after one year, because there will be lots of data in it.
Dr Stefan Rädiker
Training, Coaching, Analysis for Research & Evaluation
MAXQDA Professional Trainer
Stefan Rädiker
Posts: 254
Joined: 12.02.2007, 10:18

Return to MAXQDA in Research

Who is online

Users browsing this forum: No registered users and 1 guest