In this segment, we'll do a hands on demonstration in REDCap. Setting up a survey data collection project. We'll look at ice-cream and particularly people's perception of ice-cream in our survey project. Specifically we'll create a new survey project. We'll code an ice cream survey into that project. We'll test the survey by entering data. We'll review test data using the Export module in REDCap. We'll review test data using the Graphical Data Data View module in REDCap, and then finally we'll delete the test data setting up the project so that if we were truly doing a study, we'd be ready to go. In this simple example, we'll choose just a few variables as a proof of concept to setting up surveys. We'll collect gender much as we have before, male, female, and transgender, coded to 1, 2, and 3. This time we'll use a radio button, and it'll be a pick one option. We'll collect from th end-users or the survey recipients age in years. We'll collect that as an integer since it's type in number. We'll allow them to type in numbers between 5 and 110 years of age. We'll ask the question, do you like ice cream? We'll code that as No = 0, Yes = 1, much as we have before. Only this time we'll choose a radio button, again, as a method of entry. We'll ask them what is their favorite ice cream. In this example we'll only use three, and we'll give them options of chocolate, vanilla, and strawberry, coded in as 1, 2, and 3. And then finally we'll ask them how much do you like ice cream? This one we'll ask on a slider scale where the individual taking the survey will be asked to place the cursor on a certain line. The left hand label on that line will say, a little, and it will be coded in a 0. The right hand label will say, a lot, and it will automatically code to 100, so that wherever the individual places their cursor on that line, will code it in, it will transform it in as a number rather. Finally, we'll only ask those last two questions if they indicate on the survey question, do you like ice cream an affirmative answer, only if they answer yes. Next we'll log into REDCap. First, we'll make sure that we're logged into Coursera. At that point, we'll navigate to datacourse.org. We'll click the special Coursera login button here, and now we're in REDCap. We'll create a new project by going here. We'll name the project Example Ice Cream Survey. Just following the same directions that we did earlier. Create an empty project. We'll just build this one from scratch again. First let's enable surveys in this project then we'll go to the online designer. As before, it's creating a dummy instrument called My First Instrument. We'll click here to go in and make edits. First it starts with a record ID. We're going to leave that alone for now. First we'll code in gender. We'll make this a radio button, single answer. Now let's look at age and years. This will be a Text Box. It will validate it as an integer, a whole number. We said well it probably wouldn't make sense to have anybody below the age of five entering data and will top code it here at 110. This will validate the entry as people are putting their numbers in. Now we're looking at the question do you like ice cream? So let's add that in again as a radio button. We can choose the yes-no. It would basically do the same thing. But here, we'll stay consistent. We'll save that. Now we'll ask, what is your favorite ice cream. Here this will again be a radio button. And now we need to ask, how much so you like ice cream? Here we said we will use that visual analogue scale. It's a slider. A little. A lot. We could see here that this is going to show up as a slider bar with a label over here, a label over here. I'll ask you to trust me for now that the clicking on the bar above and dragging it will set a response and that will code the position from a 0 on the left to a 100 on the right. Let's do a bit of clean up on this name. Just to be consistent. And let's add some branching logic here. So this question only makes sense if they said yes to the ice cream question. Same here. Now let's add a new pretty labels, maybe above gender and age we'll ask, sorry we will begin a new section. Information about you. And above ice cream. Information about ice cream. Let's make each of these fields Gender. Age. And the ice cream yes or no question, required. We will not make the other two required because they are contingent. On answering this one yes. Now let's just clean up a few labels. Make them a little more human readable. And we've got a pretty good survey. We can come back here and edit as much as we want. But for now, let's go to the data collection instruments again. Well I have one instrument here I'm going to rename that I'll name it Ice Cream Reception Survey saving it and now let's enable this as a survey. Right now it's just any old data collection instrument I'm going to tell REDCap at this point I want you to render this one in survey mode so that people not logged into REDCap could take it. I'll enable it. It says, okay, you've got some other options here. First it's going to guess that the title for this survey should be the same title as the instrument. In this case that's fine. It's going to ask us if we want REDCap to number the questions for us, via auto numbering, one, two, three, four, five we'll say here no, you know we don't really want numbering there if we wanted that we would embedded in the question so all this choose custom numbered, we can choose the option of displaying all of the questions on one page or one section at a time. You'll recall that we just created two sections, one about your information, one about your ice cream information. In this case, the survey's going to be short, so we'll leave it all on one page. We could allow survey respondents to see information in aggregate after they completed the survey. There are several options here. If you're going to choose one of these options it's important to think about the minimum number of responses that you should collect before you start showing that information, mainly because of privacy and confidentiality potential issues. In this case we'll just disable it. If you wanted to show a logo Or a banner for your particular survey. You could upload a file here and it would work as you might expect. Embedding it at the top of that survey when it's going to be rendered and then finally we get down to the surveying instructions. Here, I could type anything that I want. These are the things that are going to be displayed at the top of the survey after the title. So, we will see please complete the survey below. If were doing an actual study, we would say something like the results will be used for something and something about the survey will take approximately one minute to complete. Again, anything that we wanted here it's going to be study specific, okay? Going on down, we can set if we wanted an exploration for the survey, if we wanted to stop working at a certain time. We could set things up so that survey respondents, if it were a long survey, could click a Save and Return Later button, and it would give them a code to come back and complete filling out the survey. Again, this is a short one, so we'll leave those two things alone. Once the survey finishes, then we could set it up so that the survey acknowledgement text says something that's specific for our study. Thank your taking the survey. A random drawing will be held on Saturday October 5th, 2013 to reward participants. Again, anything that we'd like, we can render there as well. At this point, let's just click on the save changes, and now we've got the information in this particular survey. Let's return back to the previous page, and we've got a survey. Over on the right you see we can always go back to those survey settings and change them. You'll also notice that there are notification options if I wanted to be notified or someone on my team to be notified every time a survey was taken we could set it up that way. Then we've even got options out here that if we were doing multiple surveys or multiple case report forms and we wanted surveys to be triggered based on input at some other time by users then we could set up that in a very complex way. This, again, is a simple one so we'll just leave the demonstration simple as well.
