WEBVTT 00:00:00.000 --> 00:00:03.992 align:middle line:90% [MUSIC PLAYING] 00:00:03.992 --> 00:00:15.970 align:middle line:90% 00:00:15.970 --> 00:00:17.890 align:middle line:84% Hello, and welcome to a short tutorial 00:00:17.890 --> 00:00:20.759 align:middle line:84% on how to calculate a one-way ANOVA using STATA. 00:00:20.759 --> 00:00:22.300 align:middle line:84% So I'm going to come up here, and I'm 00:00:22.300 --> 00:00:24.690 align:middle line:84% going to click on the Browse the Data 00:00:24.690 --> 00:00:26.840 align:middle line:84% just so that we can discuss this. 00:00:26.840 --> 00:00:28.480 align:middle line:84% In our example in the book, we had 00:00:28.480 --> 00:00:29.904 align:middle line:90% three different colored rooms. 00:00:29.904 --> 00:00:32.320 align:middle line:84% We had the white room, the yellow room, and the blue room. 00:00:32.320 --> 00:00:34.361 align:middle line:84% And then we wanted to see how long it took people 00:00:34.361 --> 00:00:36.590 align:middle line:90% to answer a 10-item quiz. 00:00:36.590 --> 00:00:38.387 align:middle line:84% And so we corresponded the time-- basically 00:00:38.387 --> 00:00:39.970 align:middle line:84% wanted to see if the color of the room 00:00:39.970 --> 00:00:44.300 align:middle line:84% impacts how quickly people are able to complete this quiz. 00:00:44.300 --> 00:00:48.270 align:middle line:84% So let's go ahead and calculate the one-way ANOVA. 00:00:48.270 --> 00:00:50.200 align:middle line:84% So we're going to come up here to Statistics. 00:00:50.200 --> 00:00:53.320 align:middle line:84% We're now going to come down to Linear models and related. 00:00:53.320 --> 00:00:56.200 align:middle line:84% We're going to click over here to ANOVA/MANOVA. 00:00:56.200 --> 00:00:58.720 align:middle line:84% We're going to click on One-way ANOVA. 00:00:58.720 --> 00:01:00.370 align:middle line:90% And we're going to click OK. 00:01:00.370 --> 00:01:04.510 align:middle line:84% For our purposes, we want the response variable to be timed, 00:01:04.510 --> 00:01:06.400 align:middle line:84% because that is the interval ratio variable. 00:01:06.400 --> 00:01:09.800 align:middle line:84% And the factor variable we want to be color. 00:01:09.800 --> 00:01:12.249 align:middle line:84% Because again, the factor one is going to be the groups. 00:01:12.249 --> 00:01:14.290 align:middle line:84% We're going to go ahead and click on the Multiple 00:01:14.290 --> 00:01:14.860 align:middle line:90% Comparisons. 00:01:14.860 --> 00:01:16.485 align:middle line:84% We're going to click Scheffe, and we're 00:01:16.485 --> 00:01:19.840 align:middle line:84% going to produce the Produce summary table. 00:01:19.840 --> 00:01:22.360 align:middle line:84% And then at that point we can just click OK. 00:01:22.360 --> 00:01:24.680 align:middle line:84% So a lot of information quickly appears. 00:01:24.680 --> 00:01:26.930 align:middle line:84% First thing you're going to see is this summary table, 00:01:26.930 --> 00:01:29.744 align:middle line:84% which is going to give you your means for each of the groups. 00:01:29.744 --> 00:01:31.660 align:middle line:84% So we have the mean and the standard deviation 00:01:31.660 --> 00:01:34.180 align:middle line:84% for the white group, for the yellow group, 00:01:34.180 --> 00:01:35.980 align:middle line:90% and for the blue group. 00:01:35.980 --> 00:01:38.710 align:middle line:84% The next is going to be the actual ANOVA a summary table. 00:01:38.710 --> 00:01:40.590 align:middle line:84% In this case, here's how you can read this. 00:01:40.590 --> 00:01:44.180 align:middle line:84% You have your degrees of freedom, which is 2 and 9. 00:01:44.180 --> 00:01:48.130 align:middle line:90% Our F value right here is 7.20. 00:01:48.130 --> 00:01:49.780 align:middle line:90% And then we have our p-value. 00:01:49.780 --> 00:01:55.120 align:middle line:84% In this case, our p-value is equal to 0.014 00:01:55.120 --> 00:01:56.680 align:middle line:90% when you round that. 00:01:56.680 --> 00:01:59.320 align:middle line:84% So this is statistically significant. 00:01:59.320 --> 00:02:01.000 align:middle line:84% And at that point, we really do not 00:02:01.000 --> 00:02:03.200 align:middle line:84% know where the statistical difference is, 00:02:03.200 --> 00:02:04.810 align:middle line:84% which is why we have to come down here 00:02:04.810 --> 00:02:07.010 align:middle line:90% to the multiple comparison test. 00:02:07.010 --> 00:02:08.560 align:middle line:84% And here's how you can understand 00:02:08.560 --> 00:02:10.270 align:middle line:90% the multiple comparison test. 00:02:10.270 --> 00:02:12.100 align:middle line:84% What it does is it looks at the yellow room 00:02:12.100 --> 00:02:13.120 align:middle line:90% with the white room. 00:02:13.120 --> 00:02:15.190 align:middle line:84% And if you look at that second number there, 00:02:15.190 --> 00:02:17.180 align:middle line:84% that is going to be our probability value. 00:02:17.180 --> 00:02:20.050 align:middle line:84% So there is no difference between the yellow room 00:02:20.050 --> 00:02:21.760 align:middle line:90% and the white room. 00:02:21.760 --> 00:02:24.400 align:middle line:84% Then we can look at the second row here, which is blue. 00:02:24.400 --> 00:02:28.130 align:middle line:84% The first one it looks at is the white room with the blue room. 00:02:28.130 --> 00:02:31.600 align:middle line:84% Now, notice the p-value there is less than 0.05, 00:02:31.600 --> 00:02:34.210 align:middle line:84% so there is a statistically significant difference 00:02:34.210 --> 00:02:36.070 align:middle line:84% between how fast it took people to take 00:02:36.070 --> 00:02:38.075 align:middle line:84% the test in the white room and the blue room. 00:02:38.075 --> 00:02:40.450 align:middle line:84% And then we can compare the blue room to the yellow room, 00:02:40.450 --> 00:02:43.510 align:middle line:84% and once again, there was not a statistically significant 00:02:43.510 --> 00:02:44.390 align:middle line:90% difference. 00:02:44.390 --> 00:02:48.530 align:middle line:84% So while the overall model was statistically significant. 00:02:48.530 --> 00:02:50.680 align:middle line:84% The real difference was between those students 00:02:50.680 --> 00:02:54.460 align:middle line:84% who took the quiz in the blue room and the white room. 00:02:54.460 --> 00:02:56.170 align:middle line:84% And that is how you can calculate 00:02:56.170 --> 00:02:59.290 align:middle line:90% a one-way ANOVA using STATA. 00:02:59.290 --> 00:03:02.940 align:middle line:90% [MUSIC PLAYING] 00:03:02.940 --> 00:03:13.258 align:middle line:90%