We would like convert a current set of linked spreadsheets into an online database. We have considered using JotForm, but are open-minded as long as data input and field names changes are easy.
If you think about a spreadsheet, it will be easier to imagine the database because it is in classic and simple rows and columns. The columns are markets (e.g. Pharmaceuticals, Oil and Gas Exploration, Database Acceleration, Human Resources). The rows are characteristics of these markets that fall into one of five categories. Categories are overall topics like "Criticality of the pain" or "Likelihood of closing deals." Each of the five categories has 3-6 rows under it. These rows are the "characteristics." Examples of characteristics that fit under the five categories are things like "Application performance is critical" under "Criticality of the pain" or "Ability to reach a decision maker" under "Likelihood of closing deals."
To be clear, a spreadsheet/database with 12 markets and 5 characteristics in each of the five categories would have 12 markets * 5 categories * 5 factors = 300 intersections -- which would be cells if still in a spreadsheet.
We have all of this in existing spreadsheets, so you will have format, behavior, formulas, and other aspects to copy.
There are two use cases for this database requiring different input forms and output displays.
Use Case 1: individual people estimate a 1-5 rating in each the 300 cells. People must be able to provide their ratings from any web-enabled location without seeing the input of other people. The system should provide auto-save during data entry, so the person doesn't lose ratings if the connection is lost in the middle...or if they want to finish later. Within a single "sheet" or input from a single person, the ratings for characteristics (rows) in a given market (column) are averaged to provide a rating for the category. Then the ratings for the categories are averaged allowing comparison and ranking between markets (columns).
We then aggregate across "sheets" (input from individuals), so basic math is done to aggregate all input into a top-level page. The math includes average (e.g. across all people providing a 1-5 rating at the intersection representing "Ability to reach a decision maker" in the market "Pharmaceutical," what is the average rating?), count (how many people rated this item?), and standard deviation (is there general agreement among people who rate this item or is there high variation?). Spreadsheets and databases typically have a standard deviation formula and whatever is included in the database should work fine.
Display for Use Case 1: should look like a one-page spreadsheet for data entry (i.e. putting in the ratings) and should look like a one-page spreadsheet for the aggregated numbers (one page for average, one page for count, and one page for standard deviation). We should be able to use color, a logo, and various notes to users of both the data entry page and the data summary page, but nothing more than you could add to empty cells in a spreadsheet. For example, a title, a logo, and space for instructions.
Use Case 2: in a *single* market, these same rows become interview questions. You can think of this as only one column with all of the rows. The interviewer will type directly into a text box representing a single characteristic (row within a column). Obviously, the input form for this use case must have larger text boxes. Boxes that expand as someone types would be nice, but scrolling may be enough. Each text box should support internal paragraph breaks. Entries in these text boxes will range from blank or 5 words up to ~200 words.
Display for Use Case 2: We will want to view interview notes in two ways: (1) all the responses from 1 interview (i.e. the same as the data input form) or (2) a subset of factors (rows) across all interviewees in a market (column), preferably a customizable subset of factors.
Feel free to send questions.
Project should be completed using a free or reasonably-priced commercial online database such as Jotform. It must be easily customizable, at least in terms of markets and characteristics by an end user with no knowledge of HTML were other technical languages. Please think of what you are creating as an online database template which will be modified when markets of interest or market characteristics change. The total number of inputs in use case one ("sheets" holding one person's input) will typically be less than 12. The total number of inputs in use case two ("individual interviews") will typically be less than 30. So the total data set size is quite small.
One more twist: it is likely this can be done in an advanced survey tool, such as Survey Gizmo or Fluid Surveys. This should NOT be a programming project. It should be a project to use an online tool to configure customizable data entry screens and then configure (printable) data display screens showing the results, including basic math to average or aggregate those results.