0:00 let's talk a little bit alright I talked
0:02 a little bit earlier about random mating
0:04 but it's a key assumption under which
0:06 you guys are operating when you
0:08 calculate with a product rule
0:10 calculation in pop stats by saying P
0:14 squared + 2 PQ you were assuming
0:16 hardy-weinberg equilibrium okay one of
0:19 the key components of that isn't random
0:21 mating so random what is random first
0:26 when someone says I chose these people
0:30 from my clinical study at random okay
0:35 I've been in a medical school for a long
0:37 time we've done a lot of medical studies
0:39 I put things up in the elevator
0:41 sometimes we want people to come in and
0:43 give us their blood okay and in a
0:48 medical school who do you think's in a
0:49 medical school medical students graduate
0:52 students anytime you go to select
0:55 anybody there's maybe some level of
0:57 criteria right like so if I was doing a
1:00 population study in diabetics you know
1:03 there's a big difference in diabetes
1:05 between African-Americans and Caucasians
1:06 right I was doing a study in diabetics
1:09 wouldn't I want the control group to be
1:12 of the same racial classification as the
1:14 test group were the question group yeah
1:17 so I'd have to impose a selection so
1:20 it's not random right it's just such a
1:22 thing as random no there isn't but the
1:30 caveat tall this is random is only with
1:33 respect to the parameter being studied
1:36 remember that because if I'm studying
1:41 certain receptor levels on the islet
1:45 cells for diabetics certain levels of
1:49 the presence of a receptor well I don't
1:52 have any preconceived knowledge of how
1:55 many receptors our cameraman is gonna
1:57 have on his eyelid cells if I select him
2:02 as a control for a study the ish the
2:07 thing I'm studying is random with
2:09 respect to that so in these genetic
2:12 studies
2:13 when you go out and you select a group
2:15 to genotype to count their alleles you
2:18 don't know their genotypes a priori you
2:22 know that they belong to some kind of a
2:23 group in the clinical study I wouldn't
2:26 know how many receptors they had on
2:27 their islet cells unless they had
2:31 diabetes so we'd have the diabetes group
2:33 and then the control group was the only
2:35 way it's random is you don't have
2:36 diabetes right and you're of a certain
2:38 racial category because we know there
2:40 are differences in the prevalence of
2:43 diabetes between the major racial
2:44 categories so remember that when the
2:47 defense attorney says to you we're these
2:50 people really selected randomly they got
2:54 them from blood banks well that's random
2:58 with respect to the STR you can pick
3:00 them that okay the other thing that
3:02 comes into random is the mating part and
3:06 I talked to you a little bit in the very
3:08 beginning of the course and this morning
3:10 about mating it's based on continental
3:13 proximity the point being is that you
3:16 don't choose your mate based on your
3:17 genotypes and it came up in our court
3:19 case and the curl and the DES redirect
3:21 by the DA was so when you chose your
3:24 wife did you know her V W a type and her
3:28 fg8 I said no I didn't and he says and
3:31 you're the one person that could have
3:33 right I said yeah I probably could have
3:35 but she wouldn't let me I said we got a
3:37 big laugh and the jury laughs and
3:39 everybody laughed well all the defense
3:42 attorney knew was the random mating he
3:45 just wanted to play around with the word
3:46 random mating that's why I'm trying to
3:49 really nail it down for you it's random
3:51 with respect to the parameter being
3:53 studied and the mating may not be random
3:56 with respect to race and geographical
3:59 proximity but it sure is with respect to
4:02 the STR genotypes right okay
