percentages, comparisons of occurences, etc

[jessebSF]

Hello all. So, I was told in the define fields forum that I was crazy to try and do what I need to do without using relationships.

What I am trying to do:

I have an evaluation database that records survey results for differrent projects, pre and post completion.

I foolishly created the survey answer fields in words, so I have now made calculation fields defining the words as numbers.

What I need to do now is the entire point of the database, and I can't figure it out.

My director is breathing down my neck to get reports of the data-

finding sets of records and doing percentages of the found set for each response.

How do I do this? I was told I need to go relational, but I don't know how, and I need this pretty quickly, with all the additional functions of percentages of found set, etc.

thanks for any help you can provide. I am an amature who got myself in way over my head, and am feeling screwed!

Thanks,

Jesse

ConvertedevaluationdatabaseCopyCopy.fp5.zip

[stanley]

Jesse:

Can you give a clear example of what exactly you need to do?

-Stanley

[jessebSF]

Stanley-

For example, I need to come up with a way for him to search all of the Surveys done Pre (Pre of Post is a field in the demographics), by project, and come up with the percentages of people who answered each question "Maybe", "Completely Disagree", "Agree Somewhat", and "Agree Fully".

I need to be able to create reports for specific projects with those percentages, etc.

Does that make sense? Thanks for trying to understand.

Jesse

[Ender]

If all the questions have the same format and choices, a relational model is a good way to go (one Survey-Answer record for each question on each Survey.)

But if you have a survey that has a variety of question formats, or a variety of choice types, then you might try the method I demonstrate in the attached files. It's a bit work to set up the fields correctly and requires a couple key scripts to populate the Question-Response counts, but can be retrofitted in an existing flat file survey easier than a relational design.

As far as narrowing the report to a specific project, you can do a find in your Project field before summarizing (my method summarizes the found set.)

Survey.zip

[jessebSF]

Holy wow, thanks Ender. Though I am still confused, your files have provided me with hope! Just to clarify though (please forgive the ignorance displayed here)- so, I have my database, with the current fields. Then, there needs to be a script in that database sending the question results (answers selected in each users record) for my found set of records (by project, or whatever else) to the other database.

Or, is this only useful if I have different answer choices? The questions I really need the reports on are the ones that all have the same answer choices.That being the case, I have a relational databse where the records in the related databse are each questions answer choices? Sorry, I am a little confused. If I have all the same answer choices and formats, I need to go relational, right? And if I do, in the related databse, what are the fields and what are the records?

Thanks so much for all your help. I REALLY appreciate it!

Jesse

[Ender]

For a relational survey (or any type of database with a master set of questions and then answers for each respondant,) you would have a file for the Survey, a file for the Questions, and a file for the Answers. The Answer file is a join between the Survey and the Questions.

So for each Survey, each required question will be a record in the Answer file. It's easiest for the user if the Answer records are all created for them, either with an import from the Question file or a loop that creates an Answer record for each Question#.

The survey could be answered through an Answer portal in Survey, but I've found if you have lots of questions, it easier for the user if the data entry is done directly in the Answer file.

The nice thing about this relational design is that reports for the responses can be made very easily with sub-summary reports in the Answer file.

[jessebSF]

Hello-

Amazingly, I am still working on this issue.

Can you just give me an example of, in this method, if I had a file for the survey, a file for the questions, and a file for the answers, what fields/data each would contain? I now understand a little more about relational databases, and I am just having a hard time figuring out, in order to do the reports I need to, which files should contain which info.

For example, currently, on each survey, there are the following elements

Organization/Project Info:

1. Pre-Project or Post-Project: (choose one)

2. Date of Survey (Insert)

3. Initiative: (choose one)

4. Project (choose one)

Respondent Info:

5. Gender

6. Age

7. Birthdate (Insert)

8. Race/Ethnicity (choose all that apply and fill in the blank if necessary)

Evaluation Questions:

21 Multiple choice (choose one), currently all with the same answer choices

5 Fill in the blank

So, which information would go where in the multiple files?

I really appreciate any help! I am bald, but if I weren't before, this process would have caused me to pull all of my hair out.

Thanks,

Jesse

P.S. I only have FMP 6

[jessebSF]

Any Help would be so appreciated on Post#178201

:qwery: :qwery: