Scores

Scores provide you with an alternative view of question results. They provide ability to define one metric that represents its underlying components in a single resulting number, making it therefore easier to digest and to compare to other scores. Instead of having to look at all of the answers, the score is a number that quickly convey a lot of information.

Scores are sometimes also referred to as “Sentiments” or “Derivative Metric” or “Index”.

Scores are defined by a mathematical formula. The formula may use the counts of each of the answer options as variables, as well as the total number of answers. Basic mathematical functions are also available: addition, subtraction, multiplication, division, exponentiation and prioritization (by grouping into parenthesis).

Example

For example, consider the question, “How much do you like to eat at Chipotle?.” This question has five answer options:

If your research objective is to track brand favorability, you may consider a score that adds the first two answer options and divides by the total number of answers. To express it in percent terms, you also want to multiply it by 100. In this way, you will have a Favorability Index that quickly conveys in one number what is the share of respondents that are favorable to Chipotle. In fact, this index is so common and important that it comes defined out of the box – see “Pre-Defined Scores” below. But you also have the ability to create and define your own scores.

Understanding the Score View

Once you have selected a score as your display option, the display changes to reflect the score view. In the left, it shows the score result at different points in time. The first one is the overall score, i.e., the score of all respondents (net of filters and weights). Then there are the results for last week, last month, last quarter and last year. For those scores, only respondents in the corresponding time period are considered. This allows for a quick understanding of how the score has evolved over time.

On the right, a simplified time view graph is shown. It contains the weekly moving average of the score over the previous 365 days. This allows for quickly understanding how the score has fluctuated over time. If a question hasn’t been asked for the full 365 previous days, then the graph will not have a score for those days.

As is the case with everything in the question results, filters and weights do apply.

Building a Score

New scores can be created from two locations.

First, for any question, you can navigate to Settings > Scores, and from there click the “Add Score” button to start building a new Score.

An even easier way to build a Score is to select the “create a new score” option conveniently placed in the Display Options menu in the header of every question.

Once you perform one of those actions, you will be put in the score creation screen.

Name

The first thing a new score is a name. This is the name that will appear in the “Display Options” menu, and will be seen by all other members of your account. We suggest that you make this name a concise description of your research goals for this score.

Example: Brand Favorability

Formula Definition

Now is time to define the formula. It does require some knowledge of math to understand how it works. You will have to enter a formula in the formula editing field, and the result of the formula will be the value of the score. There are dropdowns to help you enter your variables and functions, but you can also enter the formula directly in the formula editing field.

The Variables dropdown contains the variables that represent the response counts that can be used in the formula definition. In the Chipotle example above, “{a1}” represents the response count for the first answer option, i.e., the number of respondents that have selected the “I love it” answer option; “{a2}” represents the response count for the second answer option, and so forward; the “{total}” variable is also provided, as a short-cut to the total number of respondents that have answered that question.

Once you make a selection in the Variables dropdown, the corresponding variable is inserted in the formula editing field at the cursor position.

The Functions dropdown contains the operations that can be used with the variables: “+” for addition, “-“ for subtraction or negation, “*” for multiplication, “/” for division, “^” for exponentiation and “()” for grouping. Once you make a selection in the Functions dropdown, the corresponding operator is inserted in the formula editing field at the cursor position.

Here’s how the Favorability score looks like in the formula editing field:

Once a formula is entered, you can use the Validate button to check if the formula is syntactically correct. Clicking on the button will analyze the formula and inform of eventual syntax errors. Please see Advanced Formulas for more on Formulas.

Users then have the option to click on the “Save” button to persist the score. Once a score is saved, it is available to all users in the account.

Notice that the scores definition is inside the Settings tab. You will probably want to navigate to the Results tab to see more about the score in the many sub tabs.

Pre-Defined Scores

Some questions that follow well defined answer option patterns have pre-defined scores. These scores cannot be modified or deleted. These are the pre-defined scores and their formulas.

Favorability Index = 100* (like + love)/total

Awareness Index = 100 * (1-(neverHeard/total))

Effective Favorability Index = 100 * (like+love)/(total-neverHeard)

Modifying or deleting a score

User defined scores can be modified or deleted. To do so, first select the question for which you want to modify a score, then navigate to Settings > Scores. All user defined scores will be listed.

Click on the score you want to modify or delete.

To delete the score, click on the “Click here to delete this score” link at the bottom of the page.

To modify the score, simply modify the formula, validate it, and then save it again.

Do you still need help?

We are always available to answer your questions. For the speediest service, please email support@civicscience.com.