October 13, 200718 yr I have multiple True False tests some with over 400 questions. For the purposes of scoring them, I would like to be able to hit T or F (or 1 or 2) and then have the cursor automatically advance to the next item (i.e., field) without having to hit another key (i.e., tab or enter). Any advice on how to do that? Edited October 13, 200718 yr by Guest
October 13, 200718 yr I would think that using the mouse and choosing from radio button would work better than typing T or F . Can you explain your layout usage and User process so we can envision how it all works? BTW, are those 400 questions records or fields? It sounds like fields ... don't look now but you've got your structure wrong if so. Edited October 13, 200718 yr by Guest
October 13, 200718 yr Author Thanks for your feedback. The questions or items are fields, but the actual response to each question is stored as a record (I think). For instance, I have the following fields: FirstName, LastName, Question1, Question2, Question3,...Question 100. Is this wrong? A patient/client fills out a bubble sheet that contains his/her answers to 400 True/False questions. I then re-enter those True/False responses into filemaker. A True/False radio button sounds to be a viable solution if what I wanted in the original question isn't possible, but in the interest of being super efficient, I think it is a little quicker to just hit one of two keyboard keys and to have the cursor automatically advance to the next field. I envision this as having a small window that reads ITEM 1, then I hit the T or F key and then it reads ITEM 2 and then I hit the T or F key again, and so on. The goal: have a user enter the answers to 400 true/false questions in as few keystrokes as possible without using a mouse. Thanks ahead for feedback. It would be nice if the client/patient could enter their responses directly into the computer, but unfortunately, that isn't possible based on the proprietary rights of the tests and the way they were standardized.
October 13, 200718 yr The questions or items are fields, but the actual response to each question is stored as a record (I think). Knowing your specific setup will make all the difference in the proper answer. Can you attach the file here, provide a screenshot of your graph, or of your field definitions tab? Even a screen shot of the layout might help us. It doesn't sound like you even have a unique client ID. This is concerning if you have the answers as related records. It means that you are relating on their name (and there might be two Bill Smith's, for instance). Bottom line? We can't give ANY advice unless we know more about what you have. And be prepared for us suggesting you change your structure. :wink2:
October 13, 200718 yr You can [color:red]attach a file to your posts. When creating a new Reply, or Editing your Own Post, it is an option at the bottom of the New Reply page. Look for File Attachments: [color:red]Manage files Valid file types: zip, txt, pdf, bmp, gif, jpg, png, sit, mov Be sure to zip the file before posting it. HTH Lee p.s. the instructions are also in the FAQ (Help) above. Edited October 13, 200718 yr by Guest p.s.
October 13, 200718 yr Author Thanks Lee, It is attached here. There are some dead links to scripts as I quickly went through and deleted anything that had identifying info (e.g., letterhead, practice phone numbers and such). The basic structure of the solution is the same. I've only half way created what is needed for my original question. You have to navigate to the section called Clinical Info to see the tab for the MMPI (which is a test that I use). My_Practice.fp7.zip
October 15, 200718 yr I haven't looked at your file (too many deadlines approaching for me to spare the time) but my initial feelings are that the idea for data entry will not be a good user experience. Simply put, entering 400 answers is going to be a dull and time consuming job at the best of times. Giving the user an interface where there is no way for them to check an entry before confirming it (and thus not allowing for corrections) is going to be a major slow down. User's won't find typing Enter or Tab a problem or a slow-down, after a while they'll be doing it automatically while their eyes scan onto the next question to enter. I'd be looking at getting the people to enter the data directly into the database somehow, or implementing one of those survey processing systems (I dunno what they are called, but they have been around for decades. The ones that use the 2B pencils and multiple-choice questions.)
October 15, 200718 yr NN, I think you might like to ask your users which approach they would prefer. Are they experienced/trained data entry people or the office gofers? This could make a difference to your approach. FWIW, my own experience, on the few occasions where I've had to enter a lot of T/F type data, is that, if I make it too fast, I also make it easier to make errors. I also agree with Vaughan about the better alternative data entry approaches. Even scanning would be more productive.
October 15, 200718 yr Hi Brad, I started to re-work your file. I started to respond several times. Simply, your MMPI questions are fields (over 350 of them) in your Patients table. You don’t have a proper ClientID from which to join your tables; you are using something called Marked Relationship and Relationship. You have your ClientName also as redundant data in Points of Service. You have 14 repeaters to hold your Service Dates, start and end times in your Patients table ... I began correcting what needed to be corrected just so you could properly function. I deleted the 350+ fields and created a Questions table and Answers table and reconnected to proper IDs. But I realized that you would not even understand the changes I was making and you couldn't take it further anyway. Then I began searching for a basic structure online which you could download to show you 1) what a proper relationship looks like and 2) shows you what a good survey solution looks like. But I couldn't find one and ran out of waking hours. If anyone knows of a base structure and simple survey structure that we can provide Brad then I would appreciate the help here. :wink2:
October 15, 200718 yr I would recommend studying these two threads: http://www.fmforums.com/forum/showtopic.php?tid/180113/ http://www.fmforums.com/forum/showtopic.php?tid/183639/
October 16, 200718 yr Author Thanks for everyone's feedback. I'm reading, trying to figure out how to fix this database. It honestly seemed simple until I want to figure out a way to mark records within a portal to be exported to a word file, hence, I ended up with some redundancies. I made that part of the solution way more difficult than it needed be at that point. It was one of those things where I mimicked what I saw in somebody else's database without truly understanding how I was doing it.
October 16, 200718 yr Author Laretta, Can you please explain what you mean by "14 repeaters". I think I know what you mean as there is no reason to have redundant fields such as the patient's first and last name in the field. As far as a proper ID, doesn't my "Match" field serve that purpose (though poorly labeled, I now see). Myself and maybe two of my technicians will be the only 3 people using this solution. In psychology, our tests are supposed to be given in the exact manner in which they were standardized for the sake of the scientific method and maintain the validity of the test. I don't actually think that there would be a difference between having a patient enter scores directly into the computer versus me re-entering their bubble sheet. However, deviating from standardization is not an option with these particular tests as opposed to the bubble sheet that it was originally standardized on. That sort of deviation from standardization is something to generally avoid just in case. I'm trying to be brief without over-explaining test theory, but in short, I cannot change the patient's experience of the test. I can only make my re-entering of the data as efficient as possible.
October 17, 200718 yr There are two separate issues here. One is your data model. The other is a method for rapid data entry. The second issue is a simple one - but it needs to fit the data model. In order to be able to get meaningful statistics from your data, it MUST be organized so that each individual answer is a record (with information about which question was answered, which patient answered, and what was the answer). This means a structure with three tables: Patients, Questions and Answers - with Answers being a join table between Patients and Questions. Once you have that in place, the rapid entry problem becomes easy: for example, you could enter your answers as a string of "TFTTFTFFF...." into a temp field in Patients, then run a script that creates a record in Answers for each entered character - using the current PatientID to identify the patient, and the position of the character in the string to identify the question.
October 17, 200718 yr Here is a very simple demo showing the core relationships and a basic script for rapid data entry. Note that much more error checking is required (for example, if the number of entered characters equals the number of questions, or if all the entered characters are either T or F, etc.). RapidSurvey.fp7.zip
October 17, 200718 yr Author Thanks for the sample! I've modified to fit my needs. This should work just fine. Here is how I've modified it: resized the answer column to be just one character wide and then added text labels to denote the item I'm on for the occasion that I forget where I am. RapidSurvey.fp7.zip
Create an account or sign in to comment