Thread: Simple DB for multiple surveys
10-12-14, 08:58 #1Registered User
- Join Date
- Oct 2014
Simple DB for multiple surveys
At the company were I work, we have a survey which customers of our shop receive.
The survey consists of a part of around 5 questions which content changes only once or twice a year.
Furthermore, each two months we add a few more questions (ranging from 1-10) for a duration of 14 days. Only a percentage of our customers receive these questions.
Currently, I download the responses manually and copy-paste the non-changing questions together in an excel file. The responses on the questions that vary, are copy-pasted in a separate file. The main problem is, that Excel becomes slow and unstable at >100.000 cases. So I thought, now it is time to save the data in a database.
I neither have the time or the knowledge to set-up an efficient database design. So I was thinking about just saving all responses of one year in one table (That would be around 500.000 - 1 million responses, our customers love our surveys). So that if I want to perform an analyses, I can use a query to extract the required time frame and questions and export the responses to a excel or csv file which I can use for further analysis. What do you think about this? Is it a good or bad idea? And do you have any suggestions for a database which is simple to setup and manage? I already tried Access, but I repeatly get strange error messages in Access and it crashes regularly.
10-12-14, 10:14 #2Resident Curmudgeon
- Join Date
- Feb 2004
- In front of the computer
I see a number of choices, but don't have enough information about your needs to single out one of them over the others. Based on the way you've written your post I have the impression that you are working for a small but very quality focused business, so that adds some constraints to possible solutions.
First and foremost, if your customers are technically savvy enough to like a web page then you could hire the job done. As long as you pick a designer and a web page host that will work with you, this is probably your best option in terms of speed, ease, and general utility.
An MS-Access consultant is usually easy to find, but it can be really hard for a business person to pick a good one. If you pick a good MS-Access consultant, this can be a great solution but if you don't it can become a sink hole for both time and money.
If you spent some time in the DBForums Microsoft Access forum I'm pretty sure that you could get MS-Access to behave nicely. I get the feeling that you don't have the time to dedicate to this problem to make that a viable answer for just this problem, but if you consider the other problems that you could solve using a database then it might be a lot more attractive answer.
The short answer: If you need a finished result fast, I'd recommend using a contractor. If you need to learn to handle a database (and I'd bet that it would help you with many things), then you may need/want to invest 40-80 hours in really learning Microsoft Access.
-PatPIn theory, theory and practice are identical. In practice, theory and practice are unrelated.
10-14-14, 18:44 #3Registered User
- Join Date
- Oct 2014
You are absolutely right that I work for a small business. Of course I digitalised the survey, but I don't have direct access on the surveys database, I can only download the results as text files.
Hiring an Access consultant is a good idea and I spoke with some of them, but it's indeed not easy to find a good one. I already invested some time in learning some database and access basics (I think I invested 20-30 hours in it) but I repeatly get strange (very unspecified) error messages (Google couldn't help me further), so I thought it maybe isn't a good tool for larger datasets. I have to admit that I have limited knowledge and didn't set up any relations for example: I made one table and appended additional cases. It's probably a good idea to look around in the Access forums here, because I want to use it and expand it when it's working. I will invest more time in learning it, but currently that's not really possible.