Articles in this section

See more

How to Combine Separate Questions into a Grid in Q

  • Updated
Follow

You may have a series of Pick-Any (multiple response) questions that all share the same answer choices. Q will attempt to combine these variables into a grid, like the example below, which shows brands rated on various brand attributes (each row represents a Pick-Any question).

Requirements

  1. Completeness - there must be a variable for each row/column combination in your final table, see Additional Notes below on how to create placeholder variables if needed.
  2. Order – the order in which options appear for each attribute. You can drag and drop rows on the Variables and Questions tab line them up correctly.
  3. Labels – the labels of the questions must include both the final row and column names (Note: these are also case-sensitive). Ideally, the row/column labels should be separated by a spaces and a special character like “–”or “:” without any extra question text between or after them. Q makes it easy to edit labels by changing them directly on the Variables and Questions tab, using Search/Replace (yep, with wildcards!) or by copying/pasting using Excel.

 

 

Method

Load the data

For my example, I will use the following CSV file.

To load the data into Q:

  1. Download the file to your computer.
  2. In Q, go to File > Data Sets > Add to Project > From File, select the file, and click Open.
  3. Click OK again to have Q automatically try to group together your questions.
  4. Since we read in a CSV file, there are a lot of hidden copies of our original text variables on the Variables and Questions tab. To tidy this up and send those to the bottom, go to Automate > Browse Online Library > Move Data > Move All Hidden Questions to the Bottom.

Review the data

Let’s go to the Variables and Questions tab and inspect our Brand Attitude data (shown below). You can see the brand attitude question looking at attitudes and colas is broken into several questions.

A few things to note:

  1. Feminine is its own Pick Any question (in yellow)
  2. Health-conscious is broken into multiple Pick One questions, where the colas are out of order and labels are swapped (in blue)
  3. The rest of the attitudes are grouped into one Pick One – Multi question (in green)

Standardize the labels and order

Let’s clean up the data so we can group it together into a Pick Any – Grid.

  1. Highlight all rows of the feminine question, then right-click and select Revert to Source. This will break apart the question and show you the original labels.
  2. Manually change the labels for the health-conscious rows to look like the feminine You can also highlight the labels, copy (Ctrl + C), and paste (Ctrl + V) into Excel to modify using a formula. You can then paste back into Q.To copy:
    In excel:
    Back into Q:
  3. Drag and drop the rows on the Variables and Questions tab to make sure the options are in the same order for each set of attitudes. Reorder feminine so that None of these is at the bottom of the list of feminine options, and reorder health-conscious so that Coke comes first.

Now all your labels should look similar:

Combine the questions

Now that our labels are ordered and standardized, let’s combine them into a Pick Any – Grid question.

  1. On the Variables and Questions tab, highlight all rows of the brand attitude data.
  2. Right-click and select Set Question.
  3. Give the new question a Name like “Brand Attitude” and set the Question Type to Pick Any – Grid. Click OK.
  4. You will see a box alerting you that Q has automatically detected how to split the variable labels into row and column labels. Click OK.
  5. The Grid Layout screen that pops up gives you a preview of the row and column names, and which label corresponds to which cross-section. You may need to play with the Layout dropdown to get your grid to appear closer to what you need. Click OK.
  6. In the values window that pops up, make sure you check Missing Data for any values that suggest that the respondent hasn’t seen or answered the question. Check Count This Value (CHECK) for any response that you want to count as an affirmative response in your grid. In this example, “Yes” will be checked to Count and “No” will have no checks. Click OK.

Now you have created your own Pick Any –Grid!

 

Additional Notes

If you do not have a complete set of variables (i.e. one variable for each row/column pair), you can create dummy placeholder variables for those using missing data. To do so:

1. Go to the Variables and Questions tab.

2. Right-click on the row where you want to add the new variable and select Insert Variable(s) > JavaScript Formula > Numeric. 

3. Enter the following code into the Expression section making all value missing data:

NaN

4. Enter a Label for the new variable that matches the same pattern exactly as the other variables for the grid - including any question text at the beginning of the label.

5. Click OK to save the new variable.

6. If more placeholders are needed, you can copy and paste the new variable and modify the label as needed. 

7. Highlight all the variables you need for your grid, right click and Set Question to a Pick Any question.

8. Either use right click > Sort By > Label or drag and drop the placeholder variables into their proper place.

9. If the labeling is consistent given the example above, you can change the Question Type to Pick Any-Grid.

10. Follow from step 4 in Combine the questions section above.

 

NEXT

How to Merge Categories

How to Merge Questions in Q Without Writing Syntax

How to Merge Multiple Pick One - Multi or Grid Questions

How to Merge or NET Categories on a Pick One-Multi

 

Related articles