Part IV: Using the Results

Understanding the Results

How the EQ-i 2.0 Scores are Derived

Understanding Norms

Norms are a set of data that is collected on a large, representative sample of people. For the EQ-i 2.0, 4,000 people completed the assessment in 2010, and these responses were grouped into what we refer to as a Norm group. You may also see this Norm group referred to as the General Population, and that is essentially what it represents. Because we can’t sample everyone in the world, we take the next best thing, which is a representative group of people against which your client’s score is compared.

Norms are used as a benchmark to transform your client’s raw scores into standard scores. In the case of the EQ-i 2.0 your client’s score is compared to the scores of 4,000 people in the general population so that you know how he/she did relative to everyone else.

Scores for the EQ-i 2.0 norm group closely resemble a normal curve so you can make accurate conclusions about where the majority of respondents score on the EQ-i 2.0. Breaking the normal curve into quartiles provides a statistical reference for the 10 point cut-offs you see on the EQ-i 2.0 profile graph (i.e., low range, mid-range, high range). At the lower quartile, 25% of respondents score below 90, 50% of respondents score between 90-110 and 25% of respondents score above 110, which is the upper quartile.

Specific details on the composition of the EQ-i 2.0 Norm Group are outlined in Standardization, Reliability, and Validity and Appendix A.

Understanding Standard Scores

In the same fashion as the original EQ-i, the EQ-i 2.0 presents your client’s results by using standard scores. Standard scores are scores that have been adjusted by comparing them to others’ results on the same test. This adjustment allows you to make a comparison between scale scores (e.g., Empathy and Optimism) and makes interpretation possible because you now have a yardstick against which to measure. For instance, without standard scores it would be like your child telling you he scored a 75/100 on his math test. Would it change your opinion of his score if you knew the class average was 88/100 and that most students scored close to average? In this case, your son’s score is significantly lower than most of the rest of the class.

EQ-i 2.0 standard scores are calculated from raw scores so that each scale has the same average (mean) score of 100 and a standard deviation of 15. A standard deviation is the average or expected amount of variance in data points around the mean. A large standard deviation means that data points are far away from the mean. For example, the following set of numbers would have a large standard deviation: 0, 0, 14, 14. A small standard deviation means that data points are tight and clustered closely around the mean. For example, the following set of numbers would have a small standard deviation even though the mean is the same as above: 6, 6, 8, 8.

BENEFITS OF STANDARD SCORES

You can compare scores across different EQ-i 2.0 scales.
You can compare EQ-i 2.0 scores with other instruments standardized the same way.
You can automatically tell where the test taker’s score is, relative to the average of the normative group.

You will rarely need to explain the concept of a standard score in the terms that have been used here. Instead, your client may ask you questions such as “Is my score high/low or good/bad?” or “How does my score compare to everyone else’s?” You can answer these questions given what you now know about the EQ-i 2.0 standard scores. Your client’s score is actually generated by comparing it to the norm group. You also know that the average is 100, so depending on where your client’s result lies, you can provide an interpretation of where you client falls relative to the average.

Understanding Confidence Intervals

All measurements contain some error. Confidence Intervals1 take this error into account by providing a range of scores, at a specific level of probability, within which an individual’s true score is expected to fall. For the EQ-i 2.0 a 90% Confidence Interval was calculated, which allows you to say that 9 times out of 10 the individual’s true score would fall within the range shown. For example, your client’s Total EI score is 100. The 90% Confidence Interval for this score is 96-104 which allows you to say that nine times out of ten your client’s true score would be between 96 and 104.

Confidence Intervals can help you gauge the differences between subscales. If two confidence intervals overlap a lot for two different subscales (see example below), then an individual’s true scores on each of these subscales may not be that different from one another. If there is no overlap, or little overlap between confidence intervals, then the difference between the true scores for these subscales is probably fairly large.

Table 8.1. Example of Confidence Interval Overlap

Subscale	Score and Confidence Interval	Conclusion
Empathy	100 (95-105)	The client’s true Empathy score could fall anywhere between 95 and 105, whereas the true Flexibility score could fall anywhere between 96 and 112. Scores on these two subscales are similar.
Flexibility	104 (96-112)

Example of Confidence Intervals with no Overlap

Subscale	Score and Confidence Interval	Conclusion
Empathy	76 (71-81)	The client’s true Empathy score could fall anywhere between 71 and 81, whereas the true Flexibility score could fall anywhere between 82 and 98. Scores on these two subscales are quite different.
Flexibility	90 (82-98)