Skip to main content
What does 'Total' mean?

Everything you need to know about Totals in Knowledgehound.

Joe Razza avatar
Written by Joe Razza
Updated over a week ago

By the end of this article, you will understand what Total is, why it exists, and be able to use it and understand it when analyzing or looking at data in KnowledgeHound.

Please refer to the above Glossary while reading to understand some common terms used in this article, such as Variable, Respondent, Base, Frequency, etc.


Let's discuss what Total represents in a simple single variable case.

For example, let's say a question 'What is your Gender?' was asked to 10 respondents, 3 answered 'Male', 6 answered 'Female', and 1 did not answer at all.

Respondent

Gender

1

Male

2

Female

3

Female

4

Female

5

Male

6

Male

7

Female

8

-

9

Female

10

Female

The Total for this example would be 9 respondents. Note that the non-response, represented above as a '-', was not counted for the sum. This example reminds us that within KnowledgeHound, Total is not simply the count of the base, but rather the count of respondents who provided a response. Let's dig into this 'No Answer', or what we call *null value*, concept a little further.

Handling null values or “No Answer”

Many surveys are constructed in such a way that a respondent must answer every question. However, some surveys allow questions and/or responses to be optional or are programmed to skip questions/responses based on previous answers, or other programmatic conditions. These variations yield a null value as shown above.

The KnowledgeHound platform doesn’t currently distinguish between “did not see” and “did not answer” on single choice / pick one questions (such as the 'What is Your Gender?' example above), and treats that respondent as a non-respondent for that question.

Some systems may treat null values as “did not answer” even if the respondent did see the question, counting the respondent as part of the Base, and would yield a Total of 10 instead.

How Total works in cross tabulated single choice variable case

We introduce the question, 'What is Your Favorite Color?' as another example of a Single Choice (otherwise known as Single Select) variable. Here's how respondents answered both 'What is Your Gender?' and 'What is Your Favorite Color?'

What is your favorite color?

Gender

1

Red

Male

2

Green

Female

3

Yellow

Female

4

Blue

Female

5

Red

Male

6

Red

Male

7

Green

Female

8

Blue

-

9

Blue

Female

10

Blue

Female

Let's cross tabulate / segment these variables together and view the results as a spreadsheet.

Total

Male

Female

Blue

Frequency: 3

Base: 9

Frequency: 0

Base: 3

Frequency: 3

Base: 6

Green

Frequency: 2

Base: 9

Frequency: 0

Base: 3

Frequency: 2

Base: 6

Red

Frequency: 3

Base: 9

Frequency: 3

Base: 3

Frequency: 0

Base: 6

Yellow

Frequency: 1

Base: 9

Frequency: 0

Base: 3

Frequency: 1

Base: 6

Let's take the first row (those who answered 'Blue') as an example. The 'Total' column shows the count of unique respondents who answered 'Blue' and also provided an answer to any response in the question 'What is your Gender?'. In other words, the frequency for the 'Blue' x Total cell are the respondents who saw and answered the 'What is your Gender?' question, and selected 'Blue' for the 'What is your Favorite Color?' question.

Now wait a minute! You might be asking “Why is the Base of the Total in this cross tabulation 9 when the favorite color question has a Base of 10?”

Some tools choose to use the question being segmented as the Base. The KnowledgeHound platform (and many tools like SPSS) uses the intersection of respondents who answered all questions in the analysis as the Base, thus 9 because one respondent did not answer the Gender question. <Note: Multiple Choice, or Pick Many, questions on the other hand, may provide data that distinguishes between “did not see” and “did not answer”. KnowledgeHound detects the presence of “did not answer” data and adjusts the Base appropriately. >

Let's explore how Multiple Choice questions can change how Total is presented.

Multiple Choice / Pick Many Question

Sometimes, a respondent may be asked to pick as many attributes about a subject in question. These types of questions are known as Multiple Choice, or Pick Many, questions. To demonstrate, let’s say we asked ‘Which of the following characteristics do you care about in a Candy Bar?’ to 5 respondents and received the following results.

Person ID

Sweet

Salty

Chewy

Nutty

Crunchy

1

2

3

4

5

Which would produce a spreadsheet like shown below.

What would the Total column for this data show?

In this question, Total would be 5. It is a good reminder that Total is not simply the sum of all column values, but count of distinct respondents. And though in this case Total was equivalent to the Base, they are answering different questions - ‘Total’ describes “The count of distinct respondents that provided an answer” while the Base describes “The count of distinct respondents that saw the question”.

Pick One Grid Question

Let's work through how to interpret Total within a Pick One Grid question. A common survey question asks the respondent to answer ‘How do you feel about these products?’, and allows them to pick across multiple options. For example, you may see a form like below in a questionnaire.

I love It

Neutral

I don’t like it

Product A

Product B

Product C

These types of questions are Grid (sometimes referred to as Matrix) questions. Typically presented as a table of rows and columns like above, the respondent treats each row as a separate answer to a question (‘How do you feel about these Products?’ in this case). Notice you can pick one choice in each row (You can only checkmark either ‘I love it’, ‘Neutral’, or ‘I don’t like it’ to product A, B, C). It is an example of a Pick One Grid question.

Grid questions can be thought of as multiple questions asked in one. Though we showcase a Pick One question example here, a Grid question also can feature a Pick Many question.

Let’s see how our respondents answered in the table below.

I love it

Neutral

I don’t like it

1

Product A

Product B, Product C

2

Product A

Product B

Product C

3

Product A

Product B

Product C

4

Product B

Product A, Product C

5

Product C

Product A

Product B

6

Product A, Product B, Product C

7

Product A

Product B

Product C

8

Product A

Product B

Product C

9

Product B

Product A, Product C

10

Product A

Product B

Product C

This collection of answers would produce a spreadsheet in KnowledgeHound as shown below.

Product A

Product B

Product C

I love it

7

3

2

Neutral

3

6

3

I don’t like it

-

1

5

If we add a Total column for the products, it looks like:

Total

Product A

Product B

Product C

I love it

10

7

3

2

Neutral

9

3

6

3

I don’t like it

6

-

1

5

Let’s break down what’s going on here in a bit more detail. First, let’s define what ‘Total’ column describes within each row. Total column is describing, out of unique respondents that answered the ‘How do you feel about these Products?’ question, how many picked the option shown in the row.

  • For the ‘I love it’ row, 10 people answered ‘I love it’ for any of the choices of Product A, Product B, or Product C.

  • For the ‘Neutral’ row, 9 people answered ‘Neutral’ for any of the choices of Product A, Product B, or Product C.

  • For the ‘I don’t like it’ row, 6 people answered ‘I don’t like it’ for any of the choices of Product A, Product B, or Product C.

This example is another good reminder that Total is not simply the sum of the data intersection, but rather a count of unique respondents of an intersection. Generally, this intersection in a cross-tabulated case can be described like so: “From all unique respondents that provided an answer to ‘Question A’, how many answered ‘Option X’ in ‘Question B’?”

Summary

To wrap up, we’ll describe what Total represents in each case we presented above.

For a Single Choice question such as ‘What is your Gender?’ Total option for that variable describes “The count of unique respondents who provided an answer to the ‘What is your Gender?’ question”.

For a Pick One Grid question such as ‘How do you feel about these Products?’, we first think of the question as two questions - ‘How do you feel’ part as the responses and ‘Of these products’ as the options. If we select the Total in the options, the Total describes “From all unique respondents that provided an answer to ‘How do you feel about these Products?’, how many answered ‘I love it?’ / ‘Neutral’ / ‘I don’t like it’”.

To summarize, Total describes the count of distinct respondents of a question. You should now feel ready to interpret Totals in any shape the Analysis tool can throw at you!

Did this answer your question?