Introduction to different iQ Scoring Methods

Numeric, Single-Select and Multiselect field types can all have scoring applied within iQ; it is possible to switch off scoring for these fields if it is not desired. Text fields cannot be scored and so scoring for them is automatically excluded.

 

ALL scoring within iQ is via percentages, this is done at 3 levels (the scores roll upwards; question - section - template):

1. Question level

2. Section level

3. Template level

 

Because all scores are returned as percentages, this means you can mix and match the scoring methods throughout a single template.

 

iQ supports 4 different types of scoring, these are:

 

1. Positive/Negative (returns 0% or 100% per question, this is well suited to yes/no [or similar] type processes)

2. Weighted (this is where you have a range of values with different scores for each question, min0% max 100%)

3. Normalised Weighted (a mathematical term where weighting needs to be rescaled for some reason, meaning scores lower or greater than 0-100% can be achieved)

4. Score Modifier (a means of affecting the overall template score, useful for red flag processes, or driving tests!)

 

NB weighting can be applied to all the above scoring methods.

 

iQ Scoring – Switching Off Scoring

 

iQ Scoring – Hints & Tips

 

iQ Scoring – Controlling weighting

 

iQ Scoring – Positive/Negative

This is the simplest approach to generating a score and it is the most used. This option simply gives an output either 0 or1 (e.g.0% or 100%),because it is either on or off, this only works where you (the question set designer)want each question to provide 0% or100% outcome(excluding questions marked N/A, as these have no score).

 

The most typical application of this is where the questions elicit a “Yes” or “No” answer,however you can specify any response text you wish e.g. it could be “Compliant” and “Not Compliant”, or “OK” and “Not OK” etc.

 

It is also possible to have more than two question responses, similarly, you can use this scoring approach with multi-select; in both cases the whole question can still only result in a maximum 100% score.

 

With a positive/negative approach, all weightings within picklist responses are ignored, the score is simply calculated by the number of positive and negative responses given.

 

iQ Scoring – Positive/Negative(numeric)

Single-Select scoring =return either 100% (where a positive response was selected) or 0% (where a negative response was selected)

 

For numeric questions, where the positive/negative scoring option is selected, you must set the numeric threshold, at/above which a 100% (positive) score is returned.

 

This is controlled at the question level within the response type configuration:

 

 

iQ Scoring – Positive/Negative (single select)

Single-Select scoring =return either 100% (where a positive response was selected) or 0% (where a negative response was selected)

 

In this example, the question is “Is site area tidy and adhering to housekeeping procedures?”; the question also has the N/A response enabled.

 

 

iQ Scoring – Positive/Negative (multi select)

Multi-Select scoring = (number of positive) / (number of positive+number of negative)

 

In this example, the question is “Select all types of PPE which you must wear to prepare a sandwich”; the question also has the N/A response enabled.

 

 

iQ Scoring – Weighted Scoring

Weighting is a means of increasing or reducing the impact of question responses, the question itself or an entire section.

 

At a question level, the system looks at all the configured picklist responses for the question and takes the highest and lowest values and uses them to define the 0-100% range. In the case of numeric questions, the question set designer manually configures the range within the question.

 

 

By default, all iQ weighting is set to 1.0 (e.g.100%) which means that all answers given are unaltered from their default value.

 

Weighting can be set (as above) at a question response level and at question and section levels.

 

iQ Scoring – Weighted (numeric)

Weighting for numeric questions is controlled at the question level within the ‘configure question set ‘view.

 

 

The % range is simply driven by the range min and max values, therefore where the question weight is 1, in this example:

 

0 = 0%

2 = 40%

4 = 80%

5 = 100%

 

If the weight was changed to 2, this would not impact the question score, but would impact the score of the question, relative to its peers, within its section. However, if all questions within a given section are weighted the same way then the scores are simply averaged across that section.

 

iQ Scoring – Weighted (single select)

• This scoring method is based on potential minimum and maximum weights of available answers.
• The score can be only in range of 0% to 100%.
• The lowest available weight becomes the point of reference for 0%, and the highest for 100%.
• If only one answer is available, the score is always 100%.

 

 

iQ Scoring – Weighted (multi select)

• This scoring method is based on weights of selected responses in relation to maximum possible selected weight and minimum possible selected weight.

• The score is always in the range 0% to 100%.

• Range = max possible weight – min possible weight

 

 

iQ Scoring – Normalised Weighted

When we looked at standard weighting, we saw that the 0 – 100% range was determined by the associated picklist values, or the manually set range in the case of numeric questions.

 

With normalised weighting the question set designer can set, at a question-level, define the 0 – 100% range, overriding the picklist or numeric options offered to the user responding to the question set.

 

iQ Scoring – Normalised Weighted (numeric)

Normalised weighting applied to a numeric question facilitates an above 100% score being returned for that question, where the user responds with a permitted value that is greater than the 100% defined threshold.

 

In practical terms, this means you define the permitted range (e.g. 0-100; this is what the user can enter)and then separately define what response garners a 100% score (e.g.50) – this this example, a user response of 75 would equate to a question score of 150%.

 

This is controlled at the question level within the response type configuration:

 

 

iQ Scoring – Normalised Weighted (single select)

 

iQ Scoring – Normalised Weighted (multiselect)

 

iQ Scoring – Score Modifier

iQ score modifiers can impact the overall score dependent on one, or more, picklists values being selected within the question set where they have been configured with score modifiers.

 

Score modifiers can be positive or negative. Typically, they’re used in a “red flag, stop work” case, this might alter the overall score so dramatically that the process is deemed a fail, e.g. the threshold pass mark maybe 80%, and a score modifier might be -21%. Therefore, no matter how good, or bad, the rest of the responses are, if the user selected the “Red flag, stop work” option, the overall score will be <80%.

 

You can configure the picklist score modifier as follows:

 

 

iQ Scoring – Question Weighting & Section Scores

Section scores are calculated as an average score of applicable questions, it also includes question weight in calculation.

 

Score of a question which has weight set as 2 is twice as important as a question where weight is set as 1.

 

 

iQ Scoring – Section Weighting & Template Score

The overall template (Questionnaire)score is calculated in similar fashion to section score. The main addition is that it includes section weight in the computations.

 

Section weight is set to 1 by default, you can modify this within the Question Set configuration as follows: