1 00:00:00,000 --> 00:00:06,600 Hi, I'm Dr. Karla Lottarella at the NASA Langley Research Center. 2 00:00:06,600 --> 00:00:09,440 Hi, we're trying to solve the problem of the barking dogs. 3 00:00:09,440 --> 00:00:10,840 I might be able to help you. 4 00:00:10,840 --> 00:00:12,440 We collect a lot of data here. 5 00:00:12,440 --> 00:00:15,580 That's how we understand problems so we can solve them. 6 00:00:15,580 --> 00:00:20,120 Right now we're developing a flight simulation experiment so we can test a new display idea. 7 00:00:20,120 --> 00:00:24,160 We'll collect data from pilots to see if they can fly you and your family around more safely 8 00:00:24,160 --> 00:00:27,200 with our new display as opposed to the old one. 9 00:00:27,200 --> 00:00:29,040 We've received a lot of emails from neighborhoods. 10 00:00:29,040 --> 00:00:32,520 We really need to know the best way to keep track of all this data. 11 00:00:32,520 --> 00:00:34,320 How do you sort and collect your data? 12 00:00:34,320 --> 00:00:38,200 Well, first we start with a question or a statement of what we think is happening. 13 00:00:38,200 --> 00:00:40,640 We call that statement a hypothesis. 14 00:00:40,640 --> 00:00:44,680 So for example, in my case, I hypothesize that the new display will be better than the 15 00:00:44,680 --> 00:00:48,440 old display and that that will be true for both younger and older pilots. 16 00:00:48,440 --> 00:00:50,880 The data we collect will fill in the test matrix. 17 00:00:50,880 --> 00:00:54,000 Each data point can be considered an observation. 18 00:00:54,040 --> 00:00:59,680 In my experiment, I'm using the computer and sensors in the computer to take these observations. 19 00:00:59,680 --> 00:01:04,080 You can also take direct observations by looking at things yourself or indirect observations 20 00:01:04,080 --> 00:01:06,840 by asking other people to report things to you. 21 00:01:06,840 --> 00:01:08,760 How do we know if data is important to our problem? 22 00:01:08,760 --> 00:01:11,800 Well, hopefully your data supports your hypothesis. 23 00:01:11,800 --> 00:01:14,640 That means that the conditions you thought were important actually did influence your 24 00:01:14,640 --> 00:01:15,640 measures. 25 00:01:15,640 --> 00:01:19,800 If it doesn't, then you might want to collect some more data, some different data, or rethink 26 00:01:19,800 --> 00:01:20,800 your hypothesis. 27 00:01:20,800 --> 00:01:21,800 This is great. 28 00:01:21,800 --> 00:01:25,240 Remember the matrix Dr. D showed us in the stink problem? 29 00:01:25,240 --> 00:01:28,960 We can use that to help us rule out some of the possible sources of the problem. 30 00:01:28,960 --> 00:01:30,600 But how do we set up a matrix? 31 00:01:30,600 --> 00:01:34,880 You really should set up your test matrix ahead of time before you collect data. 32 00:01:34,880 --> 00:01:38,960 In my experiment, it's pretty simple since I only have two factors of interest, the display 33 00:01:38,960 --> 00:01:43,120 type and the age of the pilots, and I have two levels for each of these factors. 34 00:01:43,120 --> 00:01:46,800 For the display type, I have a new display and an old display, and for the age of the 35 00:01:46,800 --> 00:01:49,480 pilots, I have younger and older pilots. 36 00:01:49,480 --> 00:01:53,600 Your matrix helps you organize your data so you can look at it, analyze it, and understand 37 00:01:53,600 --> 00:01:55,520 what's important about your problem. 38 00:01:55,520 --> 00:01:58,400 We can look at the data and get an idea about the source. 39 00:01:58,400 --> 00:01:59,840 Thanks a lot for all your help. 40 00:01:59,840 --> 00:02:03,040 Now we might be able to actually make some sense of all this. 41 00:02:03,040 --> 00:02:04,040 Thanks for stopping by. 42 00:02:04,040 --> 00:02:04,540 Good luck!