WEBVTT 00:00:01.630 --> 00:00:07.260 position:50% align:middle If you're taking the NCLEX, then you need to know about CAT, or Computer Adaptive Testing. 00:00:07.260 --> 00:00:13.434 position:50% align:middle In this video, we'll not only tell you what CAT is, but also help you understand why we use it 00:00:13.434 --> 00:00:15.718 position:50% align:middle and what it means for you. 00:00:15.718 --> 00:00:20.840 position:50% align:middle The purpose of a test is to measure something like knowledge or a set of skills. 00:00:20.840 --> 00:00:26.209 position:50% align:middle Let's think of nursing knowledge, skills, and abilities as traits that can range from really low 00:00:26.209 --> 00:00:28.710 position:50% align:middle to really high along a continuum. 00:00:28.710 --> 00:00:31.780 position:50% align:middle At the high end, the student who gets As on everything. 00:00:31.780 --> 00:00:36.150 position:50% align:middle And on the other end, the student who lets just say doesn't. 00:00:36.150 --> 00:00:40.280 position:50% align:middle With the traditional test, finding out precisely where a candidate falls on that 00:00:40.280 --> 00:00:43.460 position:50% align:middle scale requires a lot of questions. 00:00:43.460 --> 00:00:47.220 position:50% align:middle That's why CAT is so important in the design of the NCLEX. 00:00:47.220 --> 00:00:52.999 position:50% align:middle Computer Adaptive Testing provides the most precise measurement with the least number of questions. 00:00:52.999 --> 00:00:58.280 position:50% align:middle The secret of CAT is adapting the questions you see based on your previous answers. 00:00:58.280 --> 00:00:59.800 position:50% align:middle Let's take a look. 00:01:01.230 --> 00:01:02.560 position:50% align:middle Here's our scale again. 00:01:02.560 --> 00:01:09.170 position:50% align:middle Let's label the scale from -3 to 3 since that's about the range we use on the NCLEX. 00:01:09.170 --> 00:01:14.670 position:50% align:middle The test begins with a question of medium difficulty, which we'll put right at 0 on our scale. 00:01:14.670 --> 00:01:19.090 position:50% align:middle The test taker can either get this question right or wrong. 00:01:19.090 --> 00:01:24.073 position:50% align:middle Suppose the candidate responds correctly, then the CAT algorithm estimates an ability greater 00:01:24.073 --> 00:01:29.850 position:50% align:middle than 0, but places a lot of uncertainty around the estimate because there's only been one question. 00:01:29.850 --> 00:01:36.210 position:50% align:middle On the graph, we might show this as a dot around 0.5 to mark our estimate, but a very wide interval 00:01:36.210 --> 00:01:39.330 position:50% align:middle to acknowledge the uncertainty of the estimate. 00:01:39.330 --> 00:01:42.560 position:50% align:middle Now the key feature of CAT is what happens next. 00:01:42.560 --> 00:01:48.480 position:50% align:middle Instead of getting a second question at random, the computer now selects a question right around 00:01:48.480 --> 00:01:50.790 position:50% align:middle the test taker's ability estimate. 00:01:50.790 --> 00:01:56.300 position:50% align:middle If the candidate answers correctly again, our estimate of the candidate's ability goes up even 00:01:56.300 --> 00:02:00.066 position:50% align:middle more since they just got an even harder question right. 00:02:00.066 --> 00:02:02.330 position:50% align:middle But something else has happened, too. 00:02:02.330 --> 00:02:05.500 position:50% align:middle The uncertainty from before has also shrunk a little. 00:02:05.500 --> 00:02:10.640 position:50% align:middle Because we have information from two questions, we have more confidence in our estimate than when the 00:02:10.640 --> 00:02:13.460 position:50% align:middle candidate had answered only one question. 00:02:13.460 --> 00:02:20.470 position:50% align:middle On a CAT exam, our estimate of a test taker's ability can change after each new question. 00:02:20.470 --> 00:02:25.693 position:50% align:middle When the test taker correctly responds, their ability estimate goes up and they see 00:02:25.693 --> 00:02:27.680 position:50% align:middle a more difficult question. 00:02:27.680 --> 00:02:32.673 position:50% align:middle When the test taker responds incorrectly, their ability estimate goes down and they 00:02:32.673 --> 00:02:34.930 position:50% align:middle see an easier question. 00:02:34.930 --> 00:02:41.100 position:50% align:middle As the candidate responds to more and more questions, the uncertainty continues to shrink and the ability 00:02:41.100 --> 00:02:44.150 position:50% align:middle estimate becomes much more precise. 00:02:44.150 --> 00:02:49.030 position:50% align:middle Let's jump ahead now and see how things might look after 85 questions. 00:02:49.030 --> 00:02:53.570 position:50% align:middle We can see that the candidate's ability estimate has increased a little, but more importantly, 00:02:53.570 --> 00:02:56.520 position:50% align:middle the uncertainty around that estimate is now much less. 00:02:56.520 --> 00:03:02.370 position:50% align:middle In fact, our ability estimate is now so precise that we may have enough information to determine whether this 00:03:02.370 --> 00:03:05.680 position:50% align:middle candidate should pass or fail the exam. 00:03:05.680 --> 00:03:10.190 position:50% align:middle For the remainder of this video, we will assume the pass/fail cutoff, 00:03:10.190 --> 00:03:14.501 position:50% align:middle sometimes known as the passing standard, is at 0.5. 00:03:14.501 --> 00:03:20.150 position:50% align:middle We can see from the graph that our candidate's ability estimate even accounting for any uncertainty is 00:03:20.150 --> 00:03:22.384 position:50% align:middle clearly above 0.5. 00:03:22.384 --> 00:03:27.540 position:50% align:middle The candidate would therefore pass the exam with what is known as a minimum length pass, 00:03:27.540 --> 00:03:32.590 position:50% align:middle 85 questions is the earliest a pass result can be determined. 00:03:32.590 --> 00:03:35.510 position:50% align:middle Of course, other results are possible as well. 00:03:35.510 --> 00:03:38.900 position:50% align:middle Let's look at a different candidate's graph after 85 questions. 00:03:38.900 --> 00:03:46.720 position:50% align:middle Here, again, the ability estimate is very precise and clearly below 0.5, even accounting for any uncertainty. 00:03:46.720 --> 00:03:51.940 position:50% align:middle This candidate fails the exam with what's known as a minimum length fail. 00:03:51.940 --> 00:03:56.220 position:50% align:middle Let's now look at a third candidate following 85 questions. 00:03:56.220 --> 00:04:02.500 position:50% align:middle Here, the ability estimate is right around 0.5, our pass/fail cutoff for this video, but the uncertainty 00:04:02.500 --> 00:04:08.440 position:50% align:middle of the estimate allows for an ability that might be just above or just below the passing standard. 00:04:08.440 --> 00:04:13.641 position:50% align:middle In other words, we do not yet know whether this candidate should pass or fail. 00:04:13.641 --> 00:04:19.217 position:50% align:middle The way we do that is to select additional questions for the candidate to answer all the way up to as many 00:04:19.217 --> 00:04:21.730 position:50% align:middle as 150 questions. 00:04:21.730 --> 00:04:27.750 position:50% align:middle By this time, the uncertainty will shrink to nearly zero and our ability estimate will be precise enough 00:04:27.750 --> 00:04:29.120 position:50% align:middle to determine a result. 00:04:29.120 --> 00:04:34.660 position:50% align:middle Sometimes the terms maximum length pass and maximum length fail are used. 00:04:34.660 --> 00:04:39.419 position:50% align:middle An ability estimate above the passing standard results in a pass and one below 00:04:39.419 --> 00:04:42.200 position:50% align:middle the passing standard results in a fail. 00:04:43.910 --> 00:04:47.980 position:50% align:middle Now that we've covered everything, here's a brief quiz. 00:04:47.980 --> 00:04:51.410 position:50% align:middle True or false, a friend's exam was the minimum length. 00:04:51.410 --> 00:04:53.900 position:50% align:middle This means they failed for sure. 00:04:56.530 --> 00:04:57.660 position:50% align:middle This is false. 00:04:57.660 --> 00:05:00.585 position:50% align:middle They could have failed, but they also could have passed. 00:05:00.585 --> 00:05:05.520 position:50% align:middle We looked at examples today of the minimum length pass and the minimum length fail. 00:05:05.520 --> 00:05:09.020 position:50% align:middle Both situations occur quite frequently on the NCLEX. 00:05:09.020 --> 00:05:16.360 position:50% align:middle Therefore, if a test taker receives the minimum number of questions, either a pass or a fail is possible. 00:05:16.360 --> 00:05:20.070 position:50% align:middle True or false, a friend's exam was the maximum length. 00:05:20.070 --> 00:05:21.700 position:50% align:middle This means they passed. 00:05:24.480 --> 00:05:26.349 position:50% align:middle This is also false. 00:05:26.349 --> 00:05:33.150 position:50% align:middle A maximum length exam occurs when the test taker's ability estimate is very close to the passing standard. 00:05:33.150 --> 00:05:35.930 position:50% align:middle If it is just below, the test taker will fail. 00:05:35.930 --> 00:05:40.080 position:50% align:middle And if it's just above, the test taker will pass. 00:05:40.080 --> 00:05:47.300 position:50% align:middle True or false, the test taker's ability estimate stays the same throughout the entire exam. 00:05:48.890 --> 00:05:51.130 position:50% align:middle This is definitely false. 00:05:51.130 --> 00:05:57.050 position:50% align:middle The whole idea of CAT is that your ability estimate changes after every question. 00:05:57.050 --> 00:06:03.300 position:50% align:middle True or false, the uncertainty of a test taker's ability estimate shrinks after every question. 00:06:05.040 --> 00:06:09.310 position:50% align:middle Yes, this is true, and that's one of the other key features of CAT. 00:06:09.310 --> 00:06:14.261 position:50% align:middle Our estimates get more and more precise after every question. 00:06:14.261 --> 00:06:20.218 position:50% align:middle The use of CAT on NCLEX ensures that your exam will take hours, not days, and will provide a 00:06:20.218 --> 00:06:22.150 position:50% align:middle very precise result. 00:06:22.150 --> 00:06:25.940 position:50% align:middle Ultimately, of course, this result will be based on how well you demonstrate 00:06:25.940 --> 00:06:29.680 position:50% align:middle your nursing knowledge, skills, and abilities. 00:06:29.680 --> 00:06:34.610 position:50% align:middle So do your best on each question and know that no matter how long your exam is, 00:06:34.610 --> 00:06:39.100 position:50% align:middle the only way to know if you passed or failed is to await your results.