With EthosCE Analytics, evaluation data can now be aggregated across your site. But if you were not using webform templates or otherwise creating consistent questions in your evaluations, you want to take this opportunity to improve your evaluation forms for 2019.
Here are some best practices for improving the quality of your aggregated data in EthosCE Analytics.
1. Always use webform templates and instruct your staff to create the evaluation from them. This will allow you to set up the questions using best practices and reduce the opportunity for mistakes. It will also ensure that you are using the same form key for components and options.
2. In general, use numbers as keys for rating questions
This example shows using numbers as keys:
Please rate your registration experience:
1|Poor
2|Average
3|Good
4|Great
5|Excellent
This will allow EthosCE Analytics to create averages per question for a course, group of courses or site-wide.
3. Try to be consistent in the scale of your rating scales. If you use a 1-10 scale on one question and 1-5 scale elsewhere, your averages won’t be accurate.
Following the directives above might result in a report like the following:.
4. Don’t use numbers as keys for non-rating questions.
This example shows a question asking an attendee’s specialty. Don’t do this:
Please select your specialty:
1|Addiction Medicine
2|Allergy/Immunology
3|Anesthesiology
4|Cardiac Electrophysiology
...
EthosCE Analytics will create an average of those responses and you end up with meaningless data in your report. You will see something like
Question Average
Please select your specialty 14.56
Obviously, that is nonsensical!
Instead, try something like this:
Please select your specialty:
addiction_medicine|Addiction Medicine
allerge_immunology|Allergy/Immunology
anesthesiology|Anesthesiology
cardiac_electrophysiology|Cardiac Electrophysiology
5. Ensure that the question text contains everything you need to see if the data is pulled without the context of seeing the printed/onscreen webform.
For example, take a question such as “Please explain why:” that references an earlier question. You might be looking at this a year from now in a standalone context and think, “Please explain why what?”
Instead, include some context. For example, “Please explain why this topic is of interest for future courses.”
It’s also possible to group questions like this under a fieldset, and EthosCE Analytics will keep the relationship, but having the context in the question is ideal.
6. When adding speaker names, be consistent. Develop a house style and stick to it.
When developing EthosCE Analytics, we observed that a speaker name might be entered different ways, for example, Dr. Jane Doe; Jane Doe MD; Jane Doe MD; Jane Doe, MD; Jane Doe, M.D., but all for the same person.
EthosCE Analytics will allow you to filter with wildcard such as “contains Jane Doe,” but in general it will be faster to see how a speaker performed across courses if the names are entered consistently.
7. Do not use numbers in your component key names.
If the numbering ever changes then the keys either 1) won’t be accurate or, if they are updated, 2) will have different key values so they can’t easily be aggregated.