WEBVTT Kind: captions Language: en 00:00:08.240 --> 00:00:10.870 align:start position:0% how<00:00:09.240> many<00:00:09.420> bluebells<00:00:09.870> are<00:00:10.110> there<00:00:10.290> in<00:00:10.380> the<00:00:10.469> wood 00:00:10.870 --> 00:00:10.880 align:start position:0% how many bluebells are there in the wood 00:00:10.880 --> 00:00:14.079 align:start position:0% how many bluebells are there in the wood there<00:00:11.880> are<00:00:12.030> millions<00:00:12.509> how<00:00:12.780> do<00:00:12.839> you<00:00:13.019> count<00:00:13.290> them 00:00:14.079 --> 00:00:14.089 align:start position:0% there are millions how do you count them 00:00:14.089 --> 00:00:19.060 align:start position:0% there are millions how do you count them well<00:00:15.089> one<00:00:15.269> way<00:00:15.480> is<00:00:15.890> to<00:00:16.890> take<00:00:17.130> one<00:00:17.730> square<00:00:18.180> meter 00:00:19.060 --> 00:00:19.070 align:start position:0% well one way is to take one square meter 00:00:19.070 --> 00:00:22.160 align:start position:0% well one way is to take one square meter count<00:00:20.070> those<00:00:20.370> carefully<00:00:21.060> and<00:00:21.810> then<00:00:21.960> multiply 00:00:22.160 --> 00:00:22.170 align:start position:0% count those carefully and then multiply 00:00:22.170 --> 00:00:25.490 align:start position:0% count those carefully and then multiply them<00:00:22.830> by<00:00:22.980> the<00:00:23.490> square<00:00:23.760> meters<00:00:24.030> of<00:00:24.390> the<00:00:24.510> wood<00:00:24.720> it 00:00:25.490 --> 00:00:25.500 align:start position:0% them by the square meters of the wood it 00:00:25.500 --> 00:00:29.060 align:start position:0% them by the square meters of the wood it would<00:00:25.560> work<00:00:25.980> if<00:00:26.910> the<00:00:27.360> sample<00:00:27.750> square<00:00:28.440> meter<00:00:28.770> we 00:00:29.060 --> 00:00:29.070 align:start position:0% would work if the sample square meter we 00:00:29.070 --> 00:00:32.569 align:start position:0% would work if the sample square meter we take<00:00:29.460> is<00:00:29.790> an<00:00:30.630> accurate<00:00:30.930> representation<00:00:31.890> of 00:00:32.569 --> 00:00:32.579 align:start position:0% take is an accurate representation of 00:00:32.579 --> 00:00:36.680 align:start position:0% take is an accurate representation of the<00:00:33.480> meter<00:00:33.690> age<00:00:33.899> of<00:00:34.079> the<00:00:34.230> word<00:00:35.629> choosing<00:00:36.629> a 00:00:36.680 --> 00:00:36.690 align:start position:0% the meter age of the word choosing a 00:00:36.690 --> 00:00:38.720 align:start position:0% the meter age of the word choosing a representative<00:00:37.410> sample<00:00:37.440> is<00:00:38.070> a<00:00:38.100> problem<00:00:38.519> for 00:00:38.720 --> 00:00:38.730 align:start position:0% representative sample is a problem for 00:00:38.730 --> 00:00:41.119 align:start position:0% representative sample is a problem for anyone<00:00:38.969> collecting<00:00:39.629> data<00:00:39.859> for<00:00:40.859> James 00:00:41.119 --> 00:00:41.129 align:start position:0% anyone collecting data for James 00:00:41.129 --> 00:00:43.250 align:start position:0% anyone collecting data for James Thompson<00:00:41.460> studying<00:00:42.449> the<00:00:42.629> effects<00:00:42.989> of<00:00:43.079> climate 00:00:43.250 --> 00:00:43.260 align:start position:0% Thompson studying the effects of climate 00:00:43.260 --> 00:00:45.380 align:start position:0% Thompson studying the effects of climate change<00:00:43.589> on<00:00:44.070> the<00:00:44.190> pollination<00:00:44.699> of<00:00:44.940> glacier 00:00:45.380 --> 00:00:45.390 align:start position:0% change on the pollination of glacier 00:00:45.390 --> 00:00:48.979 align:start position:0% change on the pollination of glacier lilies<00:00:45.600> it's<00:00:46.499> very<00:00:46.829> important<00:00:47.719> this<00:00:48.719> plant 00:00:48.979 --> 00:00:48.989 align:start position:0% lilies it's very important this plant 00:00:48.989 --> 00:00:50.329 align:start position:0% lilies it's very important this plant basically<00:00:49.230> needs<00:00:49.649> what<00:00:49.829> we<00:00:49.949> call<00:00:50.129> out 00:00:50.329 --> 00:00:50.339 align:start position:0% basically needs what we call out 00:00:50.339 --> 00:00:52.009 align:start position:0% basically needs what we call out crossing<00:00:50.879> it<00:00:50.999> needs<00:00:51.179> them<00:00:51.420> to<00:00:51.659> mate<00:00:51.989> with 00:00:52.009 --> 00:00:52.019 align:start position:0% crossing it needs them to mate with 00:00:52.019 --> 00:00:54.380 align:start position:0% crossing it needs them to mate with another<00:00:52.559> plant<00:00:53.010> and<00:00:53.339> that<00:00:53.460> involves<00:00:53.789> pollen 00:00:54.380 --> 00:00:54.390 align:start position:0% another plant and that involves pollen 00:00:54.390 --> 00:00:56.450 align:start position:0% another plant and that involves pollen being<00:00:54.600> moved<00:00:54.870> by<00:00:54.929> insects<00:00:55.499> bees<00:00:55.980> that<00:00:56.010> are 00:00:56.450 --> 00:00:56.460 align:start position:0% being moved by insects bees that are 00:00:56.460 --> 00:00:58.399 align:start position:0% being moved by insects bees that are successful<00:00:57.269> at<00:00:57.359> pollinating<00:00:57.960> these<00:00:58.140> rather 00:00:58.399 --> 00:00:58.409 align:start position:0% successful at pollinating these rather 00:00:58.409 --> 00:01:00.469 align:start position:0% successful at pollinating these rather large<00:00:58.739> flowers<00:00:59.280> are<00:00:59.489> large<00:00:59.819> animals 00:01:00.469 --> 00:01:00.479 align:start position:0% large flowers are large animals 00:01:00.479 --> 00:01:02.450 align:start position:0% large flowers are large animals specifically<00:01:01.289> there<00:01:01.559> the<00:01:01.710> Queen's<00:01:02.039> of 00:01:02.450 --> 00:01:02.460 align:start position:0% specifically there the Queen's of 00:01:02.460 --> 00:01:04.939 align:start position:0% specifically there the Queen's of bumblebees<00:01:03.179> and<00:01:03.680> bumblebees<00:01:04.680> have<00:01:04.920> a 00:01:04.939 --> 00:01:04.949 align:start position:0% bumblebees and bumblebees have a 00:01:04.949 --> 00:01:07.120 align:start position:0% bumblebees and bumblebees have a lifestyle<00:01:05.580> where<00:01:06.090> the<00:01:06.210> queen<00:01:06.450> is<00:01:06.480> the<00:01:06.840> only 00:01:07.120 --> 00:01:07.130 align:start position:0% lifestyle where the queen is the only 00:01:07.130 --> 00:01:09.860 align:start position:0% lifestyle where the queen is the only animal<00:01:08.130> that<00:01:08.160> lives<00:01:08.580> through<00:01:08.880> the<00:01:09.000> winter<00:01:09.330> so 00:01:09.860 --> 00:01:09.870 align:start position:0% animal that lives through the winter so 00:01:09.870 --> 00:01:12.170 align:start position:0% animal that lives through the winter so when<00:01:10.050> spring<00:01:10.410> comes<00:01:10.650> a<00:01:11.040> bumble<00:01:11.880> bee<00:01:11.910> queen 00:01:12.170 --> 00:01:12.180 align:start position:0% when spring comes a bumble bee queen 00:01:12.180 --> 00:01:15.110 align:start position:0% when spring comes a bumble bee queen burrows<00:01:13.020> up<00:01:13.230> from<00:01:13.440> the<00:01:13.590> soil<00:01:13.800> and<00:01:14.100> it's<00:01:14.940> her 00:01:15.110 --> 00:01:15.120 align:start position:0% burrows up from the soil and it's her 00:01:15.120 --> 00:01:16.940 align:start position:0% burrows up from the soil and it's her responsibility<00:01:15.300> to<00:01:16.230> start<00:01:16.500> a<00:01:16.590> successful 00:01:16.940 --> 00:01:16.950 align:start position:0% responsibility to start a successful 00:01:16.950 --> 00:01:20.090 align:start position:0% responsibility to start a successful nest<00:01:17.460> she<00:01:17.760> collects<00:01:18.240> pollen<00:01:18.660> and<00:01:19.050> nectar<00:01:19.100> she 00:01:20.090 --> 00:01:20.100 align:start position:0% nest she collects pollen and nectar she 00:01:20.100 --> 00:01:24.230 align:start position:0% nest she collects pollen and nectar she lays<00:01:20.310> eggs<00:01:20.640> on<00:01:20.910> a<00:01:20.940> massive<00:01:21.300> that<00:01:21.780> she<00:01:23.240> produces 00:01:24.230 --> 00:01:24.240 align:start position:0% lays eggs on a massive that she produces 00:01:24.240 --> 00:01:26.000 align:start position:0% lays eggs on a massive that she produces offspring<00:01:24.840> and<00:01:25.080> eventually<00:01:25.680> then<00:01:25.860> the 00:01:26.000 --> 00:01:26.010 align:start position:0% offspring and eventually then the 00:01:26.010 --> 00:01:27.830 align:start position:0% offspring and eventually then the offspring<00:01:26.490> become<00:01:26.790> worker<00:01:27.060> bees<00:01:27.360> who<00:01:27.600> take 00:01:27.830 --> 00:01:27.840 align:start position:0% offspring become worker bees who take 00:01:27.840 --> 00:01:30.260 align:start position:0% offspring become worker bees who take over<00:01:28.200> and<00:01:28.470> then<00:01:28.800> the<00:01:29.040> Queen<00:01:29.250> stops<00:01:29.640> foraging 00:01:30.260 --> 00:01:30.270 align:start position:0% over and then the Queen stops foraging 00:01:30.270 --> 00:01:32.990 align:start position:0% over and then the Queen stops foraging if<00:01:31.080> the<00:01:31.410> flowers<00:01:31.830> are<00:01:32.040> blooming<00:01:32.250> too<00:01:32.610> early 00:01:32.990 --> 00:01:33.000 align:start position:0% if the flowers are blooming too early 00:01:33.000 --> 00:01:37.490 align:start position:0% if the flowers are blooming too early then<00:01:33.390> these<00:01:34.080> aren't<00:01:34.380> darn<00:01:35.040> even<00:01:35.910> emerged<00:01:36.500> so 00:01:37.490 --> 00:01:37.500 align:start position:0% then these aren't darn even emerged so 00:01:37.500 --> 00:01:39.380 align:start position:0% then these aren't darn even emerged so these<00:01:37.620> guys<00:01:37.860> don't<00:01:38.250> have<00:01:38.340> any<00:01:38.880> reproduction 00:01:39.380 --> 00:01:39.390 align:start position:0% these guys don't have any reproduction 00:01:39.390 --> 00:01:42.080 align:start position:0% these guys don't have any reproduction by<00:01:39.750> seed<00:01:40.140> if<00:01:40.530> pollinators<00:01:41.430> are<00:01:41.730> out<00:01:41.880> of<00:01:42.030> the 00:01:42.080 --> 00:01:42.090 align:start position:0% by seed if pollinators are out of the 00:01:42.090 --> 00:01:44.450 align:start position:0% by seed if pollinators are out of the system<00:01:42.470> in<00:01:43.470> order<00:01:43.740> to<00:01:43.860> track<00:01:44.100> the<00:01:44.250> rate<00:01:44.430> of 00:01:44.450 --> 00:01:44.460 align:start position:0% system in order to track the rate of 00:01:44.460 --> 00:01:47.180 align:start position:0% system in order to track the rate of pollination<00:01:45.030> over<00:01:45.480> many<00:01:45.690> years<00:01:45.890> he<00:01:46.890> has<00:01:47.040> to 00:01:47.180 --> 00:01:47.190 align:start position:0% pollination over many years he has to 00:01:47.190 --> 00:01:49.490 align:start position:0% pollination over many years he has to choose<00:01:47.400> a<00:01:47.460> sample<00:01:48.090> of<00:01:48.150> flowers<00:01:48.570> that<00:01:49.110> reflects 00:01:49.490 --> 00:01:49.500 align:start position:0% choose a sample of flowers that reflects 00:01:49.500 --> 00:01:51.140 align:start position:0% choose a sample of flowers that reflects the<00:01:49.680> distribution<00:01:50.370> of<00:01:50.460> flowers<00:01:50.850> in<00:01:51.030> the 00:01:51.140 --> 00:01:51.150 align:start position:0% the distribution of flowers in the 00:01:51.150 --> 00:01:54.770 align:start position:0% the distribution of flowers in the tundra<00:01:52.100> the<00:01:53.100> classiest<00:01:53.640> statistical<00:01:54.570> way<00:01:54.720> to 00:01:54.770 --> 00:01:54.780 align:start position:0% tundra the classiest statistical way to 00:01:54.780 --> 00:01:57.050 align:start position:0% tundra the classiest statistical way to draw<00:01:55.230> a<00:01:55.440> random<00:01:55.650> sample<00:01:56.070> of<00:01:56.430> glacier<00:01:56.880> lilies 00:01:57.050 --> 00:01:57.060 align:start position:0% draw a random sample of glacier lilies 00:01:57.060 --> 00:01:59.390 align:start position:0% draw a random sample of glacier lilies would<00:01:57.660> be<00:01:57.840> to<00:01:58.050> go<00:01:58.230> and<00:01:58.500> enumerate<00:01:58.830> each 00:01:59.390 --> 00:01:59.400 align:start position:0% would be to go and enumerate each 00:01:59.400 --> 00:02:01.900 align:start position:0% would be to go and enumerate each glacier<00:01:59.850> lily<00:02:00.120> in<00:02:00.420> the<00:02:00.630> in<00:02:00.810> the<00:02:01.110> field<00:02:01.380> here 00:02:01.900 --> 00:02:01.910 align:start position:0% glacier lily in the in the field here 00:02:01.910 --> 00:02:04.850 align:start position:0% glacier lily in the in the field here give<00:02:02.910> each<00:02:03.060> one<00:02:03.330> an<00:02:03.570> individual<00:02:04.230> code<00:02:04.530> and 00:02:04.850 --> 00:02:04.860 align:start position:0% give each one an individual code and 00:02:04.860 --> 00:02:07.160 align:start position:0% give each one an individual code and then<00:02:05.400> use<00:02:05.610> a<00:02:05.640> computer<00:02:06.090> and<00:02:06.240> a<00:02:06.450> random<00:02:06.810> numbers 00:02:07.160 --> 00:02:07.170 align:start position:0% then use a computer and a random numbers 00:02:07.170 --> 00:02:09.709 align:start position:0% then use a computer and a random numbers generator<00:02:07.710> to<00:02:08.310> pick<00:02:08.519> a<00:02:08.550> random<00:02:08.879> set<00:02:09.360> of<00:02:09.420> those 00:02:09.709 --> 00:02:09.719 align:start position:0% generator to pick a random set of those 00:02:09.719 --> 00:02:12.440 align:start position:0% generator to pick a random set of those and<00:02:10.019> go<00:02:10.140> do<00:02:10.319> them<00:02:10.499> I<00:02:10.679> can't<00:02:11.069> do<00:02:11.219> that<00:02:11.299> they'd<00:02:12.299> be 00:02:12.440 --> 00:02:12.450 align:start position:0% and go do them I can't do that they'd be 00:02:12.450 --> 00:02:14.750 align:start position:0% and go do them I can't do that they'd be out<00:02:12.599> of<00:02:12.629> balloon<00:02:13.019> before<00:02:13.530> I<00:02:13.560> done<00:02:13.950> it<00:02:14.099> so<00:02:14.700> I 00:02:14.750 --> 00:02:14.760 align:start position:0% out of balloon before I done it so I 00:02:14.760 --> 00:02:15.590 align:start position:0% out of balloon before I done it so I need<00:02:15.150> to 00:02:15.590 --> 00:02:15.600 align:start position:0% need to 00:02:15.600 --> 00:02:17.030 align:start position:0% need to take<00:02:15.930> a<00:02:15.960> sample<00:02:16.440> that's<00:02:16.590> going<00:02:16.860> to<00:02:16.950> be 00:02:17.030 --> 00:02:17.040 align:start position:0% take a sample that's going to be 00:02:17.040 --> 00:02:20.810 align:start position:0% take a sample that's going to be representative<00:02:17.880> I<00:02:18.090> try<00:02:18.630> to<00:02:18.660> spread<00:02:19.530> each<00:02:20.460> of 00:02:20.810 --> 00:02:20.820 align:start position:0% representative I try to spread each of 00:02:20.820 --> 00:02:23.780 align:start position:0% representative I try to spread each of my<00:02:21.560> selections<00:02:22.560> over<00:02:23.100> a<00:02:23.130> large<00:02:23.520> enough 00:02:23.780 --> 00:02:23.790 align:start position:0% my selections over a large enough 00:02:23.790 --> 00:02:26.630 align:start position:0% my selections over a large enough spatial<00:02:24.360> area<00:02:24.720> that<00:02:25.080> I<00:02:25.110> won't<00:02:25.440> be<00:02:25.740> getting 00:02:26.630 --> 00:02:26.640 align:start position:0% spatial area that I won't be getting 00:02:26.640 --> 00:02:29.090 align:start position:0% spatial area that I won't be getting strange<00:02:27.120> results<00:02:27.660> just<00:02:27.870> because<00:02:28.170> something 00:02:29.090 --> 00:02:29.100 align:start position:0% strange results just because something 00:02:29.100 --> 00:02:31.460 align:start position:0% strange results just because something weird<00:02:29.340> happened<00:02:29.880> right<00:02:30.090> here<00:02:30.120> or<00:02:30.810> some<00:02:31.230> deer 00:02:31.460 --> 00:02:31.470 align:start position:0% weird happened right here or some deer 00:02:31.470 --> 00:02:33.920 align:start position:0% weird happened right here or some deer came<00:02:31.740> and<00:02:32.010> started<00:02:32.220> grazing<00:02:32.940> and<00:02:33.150> took<00:02:33.630> all<00:02:33.780> my 00:02:33.920 --> 00:02:33.930 align:start position:0% came and started grazing and took all my 00:02:33.930 --> 00:02:36.620 align:start position:0% came and started grazing and took all my plants<00:02:34.680> so<00:02:35.370> I<00:02:35.400> have<00:02:35.640> to<00:02:35.760> pay<00:02:35.910> attention<00:02:36.030> to 00:02:36.620 --> 00:02:36.630 align:start position:0% plants so I have to pay attention to 00:02:36.630 --> 00:02:39.170 align:start position:0% plants so I have to pay attention to getting<00:02:37.490> representative<00:02:38.490> plants<00:02:38.880> I'd<00:02:39.030> say 00:02:39.170 --> 00:02:39.180 align:start position:0% getting representative plants I'd say 00:02:39.180 --> 00:02:41.510 align:start position:0% getting representative plants I'd say that's<00:02:39.330> the<00:02:39.570> most<00:02:39.690> important<00:02:40.230> thing<00:02:40.400> so<00:02:41.400> in 00:02:41.510 --> 00:02:41.520 align:start position:0% that's the most important thing so in 00:02:41.520 --> 00:02:43.010 align:start position:0% that's the most important thing so in this<00:02:41.610> film<00:02:41.880> we<00:02:41.940> look<00:02:42.180> at<00:02:42.300> sampling<00:02:42.840> and 00:02:43.010 --> 00:02:43.020 align:start position:0% this film we look at sampling and 00:02:43.020 --> 00:02:46.850 align:start position:0% this film we look at sampling and accuracy<00:02:43.770> how<00:02:44.310> big<00:02:44.580> a<00:02:44.610> sample<00:02:45.300> do<00:02:45.480> we<00:02:45.600> need<00:02:45.930> to 00:02:46.850 --> 00:02:46.860 align:start position:0% accuracy how big a sample do we need to 00:02:46.860 --> 00:02:49.310 align:start position:0% accuracy how big a sample do we need to be<00:02:46.980> able<00:02:47.130> to<00:02:47.430> make<00:02:47.460> conclusions<00:02:47.940> within<00:02:48.900> some 00:02:49.310 --> 00:02:49.320 align:start position:0% be able to make conclusions within some 00:02:49.320 --> 00:02:52.910 align:start position:0% be able to make conclusions within some known<00:02:49.590> margin<00:02:50.190> of<00:02:50.310> error<00:02:51.200> this<00:02:52.200> is<00:02:52.350> important 00:02:52.910 --> 00:02:52.920 align:start position:0% known margin of error this is important 00:02:52.920 --> 00:02:54.760 align:start position:0% known margin of error this is important whether<00:02:53.130> you're<00:02:53.370> studying<00:02:53.730> wastewater 00:02:54.760 --> 00:02:54.770 align:start position:0% whether you're studying wastewater 00:02:54.770 --> 00:02:59.450 align:start position:0% whether you're studying wastewater climate<00:02:55.770> change<00:02:56.100> or<00:02:56.420> politics<00:02:58.010> in<00:02:59.010> an<00:02:59.160> earlier 00:02:59.450 --> 00:02:59.460 align:start position:0% climate change or politics in an earlier 00:02:59.460 --> 00:03:02.150 align:start position:0% climate change or politics in an earlier film<00:02:59.730> we<00:03:00.300> followed<00:03:00.630> Caroline<00:03:01.140> wooldridge<00:03:01.620> as 00:03:02.150 --> 00:03:02.160 align:start position:0% film we followed Caroline wooldridge as 00:03:02.160 --> 00:03:03.560 align:start position:0% film we followed Caroline wooldridge as she<00:03:02.340> campaigned<00:03:02.760> for<00:03:02.910> a<00:03:03.030> seat<00:03:03.300> on<00:03:03.330> the 00:03:03.560 --> 00:03:03.570 align:start position:0% she campaigned for a seat on the 00:03:03.570 --> 00:03:06.560 align:start position:0% she campaigned for a seat on the Basingstoke<00:03:04.140> County<00:03:04.590> Council<00:03:05.360> the<00:03:06.360> labor 00:03:06.560 --> 00:03:06.570 align:start position:0% Basingstoke County Council the labor 00:03:06.570 --> 00:03:08.150 align:start position:0% Basingstoke County Council the labor party<00:03:06.840> here<00:03:07.290> in<00:03:07.320> Brighton<00:03:07.830> hill<00:03:08.010> in 00:03:08.150 --> 00:03:08.160 align:start position:0% party here in Brighton hill in 00:03:08.160 --> 00:03:10.850 align:start position:0% party here in Brighton hill in basingstoke<00:03:08.760> was<00:03:09.330> trying<00:03:09.600> to<00:03:09.690> decide<00:03:10.050> whether 00:03:10.850 --> 00:03:10.860 align:start position:0% basingstoke was trying to decide whether 00:03:10.860 --> 00:03:12.440 align:start position:0% basingstoke was trying to decide whether they<00:03:11.070> might<00:03:11.220> be<00:03:11.370> able<00:03:11.520> to<00:03:11.730> win<00:03:11.880> a<00:03:11.940> place<00:03:12.210> on<00:03:12.420> the 00:03:12.440 --> 00:03:12.450 align:start position:0% they might be able to win a place on the 00:03:12.450 --> 00:03:14.420 align:start position:0% they might be able to win a place on the council<00:03:12.660> that's<00:03:13.470> tended<00:03:13.920> to<00:03:14.010> be<00:03:14.160> liberal 00:03:14.420 --> 00:03:14.430 align:start position:0% council that's tended to be liberal 00:03:14.430 --> 00:03:16.940 align:start position:0% council that's tended to be liberal democrats<00:03:15.150> how<00:03:16.140> can<00:03:16.320> they<00:03:16.410> know<00:03:16.590> if<00:03:16.710> they<00:03:16.830> have 00:03:16.940 --> 00:03:16.950 align:start position:0% democrats how can they know if they have 00:03:16.950 --> 00:03:19.910 align:start position:0% democrats how can they know if they have a<00:03:17.010> chance<00:03:17.280> to<00:03:17.550> win<00:03:18.020> should<00:03:19.020> they<00:03:19.140> invest<00:03:19.500> extra 00:03:19.910 --> 00:03:19.920 align:start position:0% a chance to win should they invest extra 00:03:19.920 --> 00:03:22.330 align:start position:0% a chance to win should they invest extra time<00:03:20.160> and<00:03:20.430> energy<00:03:20.790> if<00:03:21.060> they<00:03:21.390> have<00:03:21.510> a<00:03:21.570> chance 00:03:22.330 --> 00:03:22.340 align:start position:0% time and energy if they have a chance 00:03:22.340 --> 00:03:24.940 align:start position:0% time and energy if they have a chance brighton<00:03:23.340> hell<00:03:23.610> is<00:03:23.850> interesting<00:03:24.480> because 00:03:24.940 --> 00:03:24.950 align:start position:0% brighton hell is interesting because 00:03:24.950 --> 00:03:30.710 align:start position:0% brighton hell is interesting because it's<00:03:25.950> been<00:03:26.640> clearly<00:03:27.350> Liberal<00:03:28.350> Democrats<00:03:29.720> but 00:03:30.710 --> 00:03:30.720 align:start position:0% it's been clearly Liberal Democrats but 00:03:30.720 --> 00:03:33.710 align:start position:0% it's been clearly Liberal Democrats but Carolyn<00:03:31.380> feels<00:03:31.770> she's<00:03:32.520> got<00:03:32.730> a<00:03:32.760> genuine<00:03:33.360> chance 00:03:33.710 --> 00:03:33.720 align:start position:0% Carolyn feels she's got a genuine chance 00:03:33.720 --> 00:03:36.590 align:start position:0% Carolyn feels she's got a genuine chance this<00:03:34.110> time<00:03:34.410> yeah<00:03:35.100> that's<00:03:35.280> realistic<00:03:35.640> that's 00:03:36.590 --> 00:03:36.600 align:start position:0% this time yeah that's realistic that's 00:03:36.600 --> 00:03:39.620 align:start position:0% this time yeah that's realistic that's good<00:03:37.020> I<00:03:37.230> mean<00:03:37.800> amazing<00:03:38.640> if<00:03:38.730> you<00:03:38.820> change<00:03:39.090> it<00:03:39.270> but 00:03:39.620 --> 00:03:39.630 align:start position:0% good I mean amazing if you change it but 00:03:39.630 --> 00:03:43.370 align:start position:0% good I mean amazing if you change it but she's<00:03:39.810> worked<00:03:40.050> very<00:03:41.040> hard<00:03:41.160> over<00:03:41.780> very<00:03:42.780> over 00:03:43.370 --> 00:03:43.380 align:start position:0% she's worked very hard over very over 00:03:43.380 --> 00:03:45.350 align:start position:0% she's worked very hard over very over this<00:03:43.500> period<00:03:43.800> would<00:03:44.430> think<00:03:45.120> that's 00:03:45.350 --> 00:03:45.360 align:start position:0% this period would think that's 00:03:45.360 --> 00:03:47.660 align:start position:0% this period would think that's translating<00:03:46.050> into<00:03:46.170> vote<00:03:46.440> it's<00:03:47.130> big<00:03:47.310> swing<00:03:47.520> to 00:03:47.660 --> 00:03:47.670 align:start position:0% translating into vote it's big swing to 00:03:47.670 --> 00:03:49.430 align:start position:0% translating into vote it's big swing to win<00:03:47.790> if<00:03:48.090> you<00:03:48.120> don't<00:03:48.240> see<00:03:48.780> my<00:03:48.900> swing<00:03:49.230> what 00:03:49.430 --> 00:03:49.440 align:start position:0% win if you don't see my swing what 00:03:49.440 --> 00:03:50.990 align:start position:0% win if you don't see my swing what everything<00:03:49.860> you<00:03:49.980> said<00:03:50.220> you're<00:03:50.580> talking 00:03:50.990 --> 00:03:51.000 align:start position:0% everything you said you're talking 00:03:51.000 --> 00:03:52.730 align:start position:0% everything you said you're talking you're<00:03:51.240> talking<00:03:51.300> easily<00:03:52.050> plus<00:03:52.320> twenty-five 00:03:52.730 --> 00:03:52.740 align:start position:0% you're talking easily plus twenty-five 00:03:52.740 --> 00:03:54.920 align:start position:0% you're talking easily plus twenty-five percent<00:03:52.950> TLC<00:03:53.790> I<00:03:53.910> mean<00:03:54.090> that's<00:03:54.210> that's<00:03:54.690> out 00:03:54.920 --> 00:03:54.930 align:start position:0% percent TLC I mean that's that's out 00:03:54.930 --> 00:03:56.570 align:start position:0% percent TLC I mean that's that's out there<00:03:55.200> I<00:03:55.230> mean<00:03:55.440> that's<00:03:55.530> right<00:03:55.920> on<00:03:56.130> the<00:03:56.160> edge<00:03:56.370> of 00:03:56.570 --> 00:03:56.580 align:start position:0% there I mean that's right on the edge of 00:03:56.580 --> 00:03:59.420 align:start position:0% there I mean that's right on the edge of probability<00:03:57.240> yes<00:03:57.750> but<00:03:58.170> at<00:03:58.260> the<00:03:58.380> same<00:03:58.590> time<00:03:58.980> the 00:03:59.420 --> 00:03:59.430 align:start position:0% probability yes but at the same time the 00:03:59.430 --> 00:04:00.830 align:start position:0% probability yes but at the same time the feedback<00:03:59.760> on<00:03:59.910> the<00:03:59.970> doorstep<00:04:00.360> is<00:04:00.540> this 00:04:00.830 --> 00:04:00.840 align:start position:0% feedback on the doorstep is this 00:04:00.840 --> 00:04:02.720 align:start position:0% feedback on the doorstep is this possible<00:04:01.080> encouraged<00:04:02.010> by<00:04:02.130> positive 00:04:02.720 --> 00:04:02.730 align:start position:0% possible encouraged by positive 00:04:02.730 --> 00:04:05.270 align:start position:0% possible encouraged by positive responses<00:04:03.270> on<00:04:03.450> the<00:04:03.540> doorstep<00:04:04.050> at<00:04:04.650> national 00:04:05.270 --> 00:04:05.280 align:start position:0% responses on the doorstep at national 00:04:05.280 --> 00:04:07.520 align:start position:0% responses on the doorstep at national opinion<00:04:05.550> polls<00:04:05.970> that<00:04:06.540> indicated<00:04:07.170> a<00:04:07.260> swing 00:04:07.520 --> 00:04:07.530 align:start position:0% opinion polls that indicated a swing 00:04:07.530 --> 00:04:10.130 align:start position:0% opinion polls that indicated a swing away<00:04:07.650> from<00:04:08.070> the<00:04:08.190> liberal<00:04:08.340> democrats<00:04:09.140> labor 00:04:10.130 --> 00:04:10.140 align:start position:0% away from the liberal democrats labor 00:04:10.140 --> 00:04:12.740 align:start position:0% away from the liberal democrats labor was<00:04:10.410> mobilized<00:04:10.950> in<00:04:11.190> basingstoke<00:04:11.850> to<00:04:12.510> work 00:04:12.740 --> 00:04:12.750 align:start position:0% was mobilized in basingstoke to work 00:04:12.750 --> 00:04:15.290 align:start position:0% was mobilized in basingstoke to work hard<00:04:13.200> to<00:04:13.410> take<00:04:13.770> the<00:04:13.950> seat<00:04:14.220> one<00:04:14.850> reason<00:04:15.150> why<00:04:15.270> I 00:04:15.290 --> 00:04:15.300 align:start position:0% hard to take the seat one reason why I 00:04:15.300 --> 00:04:16.400 align:start position:0% hard to take the seat one reason why I think<00:04:15.480> we<00:04:15.570> stand<00:04:15.750> a<00:04:15.870> chance<00:04:15.900> this<00:04:16.230> time 00:04:16.400 --> 00:04:16.410 align:start position:0% think we stand a chance this time 00:04:16.410 --> 00:04:17.990 align:start position:0% think we stand a chance this time because<00:04:16.680> people<00:04:17.130> are<00:04:17.190> just<00:04:17.340> welcoming<00:04:17.850> us 00:04:17.990 --> 00:04:18.000 align:start position:0% because people are just welcoming us 00:04:18.000 --> 00:04:21.140 align:start position:0% because people are just welcoming us back<00:04:18.030> so<00:04:18.690> and<00:04:19.380> it<00:04:19.680> has<00:04:19.890> been<00:04:20.130> swinging<00:04:20.730> our<00:04:20.970> way 00:04:21.140 --> 00:04:21.150 align:start position:0% back so and it has been swinging our way 00:04:21.150 --> 00:04:22.910 align:start position:0% back so and it has been swinging our way certainly<00:04:21.660> they<00:04:21.989> last<00:04:22.200> especially<00:04:22.710> the<00:04:22.800> last 00:04:22.910 --> 00:04:22.920 align:start position:0% certainly they last especially the last 00:04:22.920 --> 00:04:24.500 align:start position:0% certainly they last especially the last few<00:04:23.190> nights<00:04:23.370> it's<00:04:23.640> been<00:04:23.820> really<00:04:24.000> swinging<00:04:24.360> our 00:04:24.500 --> 00:04:24.510 align:start position:0% few nights it's been really swinging our 00:04:24.510 --> 00:04:25.170 align:start position:0% few nights it's been really swinging our ways 00:04:25.170 --> 00:04:25.180 align:start position:0% ways 00:04:25.180 --> 00:04:27.640 align:start position:0% ways one<00:04:26.180> important<00:04:26.660> way<00:04:26.720> to<00:04:26.870> influence<00:04:27.440> the 00:04:27.640 --> 00:04:27.650 align:start position:0% one important way to influence the 00:04:27.650 --> 00:04:32.620 align:start position:0% one important way to influence the outcome<00:04:28.100> is<00:04:28.310> by<00:04:28.820> using<00:04:29.060> tellers<00:04:31.150> each<00:04:32.150> party 00:04:32.620 --> 00:04:32.630 align:start position:0% outcome is by using tellers each party 00:04:32.630 --> 00:04:34.900 align:start position:0% outcome is by using tellers each party keeps<00:04:33.230> lists<00:04:33.650> of<00:04:33.740> known<00:04:34.010> supporters<00:04:34.730> and 00:04:34.900 --> 00:04:34.910 align:start position:0% keeps lists of known supporters and 00:04:34.910 --> 00:04:37.300 align:start position:0% keeps lists of known supporters and tellers<00:04:35.720> keep<00:04:35.960> track<00:04:36.230> of<00:04:36.260> who's<00:04:36.590> voted<00:04:37.070> in<00:04:37.220> the 00:04:37.300 --> 00:04:37.310 align:start position:0% tellers keep track of who's voted in the 00:04:37.310 --> 00:04:40.900 align:start position:0% tellers keep track of who's voted in the who<00:04:37.430> hasn't<00:04:38.050> this<00:04:39.050> is<00:04:39.170> looking<00:04:39.560> up<00:04:39.860> like<00:04:40.550> what 00:04:40.900 --> 00:04:40.910 align:start position:0% who hasn't this is looking up like what 00:04:40.910 --> 00:04:42.940 align:start position:0% who hasn't this is looking up like what were<00:04:41.090> you<00:04:41.180> have<00:04:41.330> to<00:04:41.450> do<00:04:41.600> now<00:04:41.810> is<00:04:41.840> you<00:04:42.740> have<00:04:42.920> a 00:04:42.940 --> 00:04:42.950 align:start position:0% were you have to do now is you have a 00:04:42.950 --> 00:04:44.830 align:start position:0% were you have to do now is you have a list<00:04:43.220> here<00:04:43.340> of<00:04:43.640> these<00:04:44.180> are<00:04:44.300> all<00:04:44.420> people<00:04:44.630> who 00:04:44.830 --> 00:04:44.840 align:start position:0% list here of these are all people who 00:04:44.840 --> 00:04:46.270 align:start position:0% list here of these are all people who where<00:04:45.080> they've<00:04:45.230> been<00:04:45.260> canvassed<00:04:45.890> but<00:04:46.100> on<00:04:46.190> the 00:04:46.270 --> 00:04:46.280 align:start position:0% where they've been canvassed but on the 00:04:46.280 --> 00:04:48.430 align:start position:0% where they've been canvassed but on the telephone<00:04:46.600> they've<00:04:47.600> pledged<00:04:48.020> that<00:04:48.290> they're 00:04:48.430 --> 00:04:48.440 align:start position:0% telephone they've pledged that they're 00:04:48.440 --> 00:04:51.610 align:start position:0% telephone they've pledged that they're going<00:04:48.560> to<00:04:48.620> vote<00:04:48.800> Labour<00:04:49.070> right<00:04:49.820> and<00:04:50.350> we<00:04:51.350> cross 00:04:51.610 --> 00:04:51.620 align:start position:0% going to vote Labour right and we cross 00:04:51.620 --> 00:04:54.040 align:start position:0% going to vote Labour right and we cross their<00:04:51.950> names<00:04:52.310> off<00:04:52.580> as<00:04:52.850> the<00:04:53.240> laser<00:04:53.450> boats<00:04:53.780> come 00:04:54.040 --> 00:04:54.050 align:start position:0% their names off as the laser boats come 00:04:54.050 --> 00:04:55.720 align:start position:0% their names off as the laser boats come in<00:04:54.230> which<00:04:54.560> we<00:04:54.710> can<00:04:54.830> tell<00:04:55.010> them<00:04:55.250> the<00:04:55.340> polling 00:04:55.720 --> 00:04:55.730 align:start position:0% in which we can tell them the polling 00:04:55.730 --> 00:04:57.760 align:start position:0% in which we can tell them the polling number<00:04:56.120> right<00:04:56.720> so<00:04:56.930> what<00:04:57.080> we<00:04:57.170> have<00:04:57.350> here<00:04:57.380> is<00:04:57.590> a 00:04:57.760 --> 00:04:57.770 align:start position:0% number right so what we have here is a 00:04:57.770 --> 00:05:00.190 align:start position:0% number right so what we have here is a list<00:04:58.040> of<00:04:58.190> people<00:04:58.490> who<00:04:58.550> haven't<00:04:58.760> yet<00:04:59.000> voted<00:04:59.240> so 00:05:00.190 --> 00:05:00.200 align:start position:0% list of people who haven't yet voted so 00:05:00.200 --> 00:05:01.270 align:start position:0% list of people who haven't yet voted so what<00:05:00.350> we're<00:05:00.500> doing<00:05:00.530> late<00:05:00.890> in<00:05:01.070> the<00:05:01.160> afternoon 00:05:01.270 --> 00:05:01.280 align:start position:0% what we're doing late in the afternoon 00:05:01.280 --> 00:05:02.950 align:start position:0% what we're doing late in the afternoon we're<00:05:01.820> actually<00:05:01.940> going<00:05:02.210> out<00:05:02.420> and<00:05:02.600> knocking<00:05:02.690> on 00:05:02.950 --> 00:05:02.960 align:start position:0% we're actually going out and knocking on 00:05:02.960 --> 00:05:05.500 align:start position:0% we're actually going out and knocking on the<00:05:03.080> door<00:05:03.290> and<00:05:03.920> saying<00:05:04.160> excuse<00:05:04.550> me<00:05:04.790> Griffin 00:05:05.500 --> 00:05:05.510 align:start position:0% the door and saying excuse me Griffin 00:05:05.510 --> 00:05:07.960 align:start position:0% the door and saying excuse me Griffin the<00:05:05.600> local<00:05:05.870> labour<00:05:06.020> party<00:05:06.500> we<00:05:07.280> have<00:05:07.490> recorded 00:05:07.960 --> 00:05:07.970 align:start position:0% the local labour party we have recorded 00:05:07.970 --> 00:05:09.190 align:start position:0% the local labour party we have recorded here<00:05:08.120> that<00:05:08.330> someone's<00:05:08.630> promised<00:05:08.930> to<00:05:09.050> vote 00:05:09.190 --> 00:05:09.200 align:start position:0% here that someone's promised to vote 00:05:09.200 --> 00:05:11.050 align:start position:0% here that someone's promised to vote Labour<00:05:09.440> and<00:05:09.620> you<00:05:09.740> haven't<00:05:10.010> voted<00:05:10.340> yet<00:05:10.580> would 00:05:11.050 --> 00:05:11.060 align:start position:0% Labour and you haven't voted yet would 00:05:11.060 --> 00:05:14.340 align:start position:0% Labour and you haven't voted yet would you<00:05:11.150> like<00:05:11.300> to<00:05:11.390> get<00:05:11.540> out<00:05:11.630> and<00:05:11.780> vote<00:05:11.840> right<00:05:12.650> ok 00:05:14.340 --> 00:05:14.350 align:start position:0% you like to get out and vote right ok 00:05:14.350 --> 00:05:17.110 align:start position:0% you like to get out and vote right ok hello<00:05:15.350> there<00:05:15.560> look<00:05:16.010> Labour<00:05:16.280> Party<00:05:16.790> have<00:05:17.030> you 00:05:17.110 --> 00:05:17.120 align:start position:0% hello there look Labour Party have you 00:05:17.120 --> 00:05:19.480 align:start position:0% hello there look Labour Party have you bolted<00:05:17.510> you<00:05:17.660> have<00:05:18.650> you<00:05:18.740> been<00:05:18.890> out<00:05:19.040> there<00:05:19.190> bones 00:05:19.480 --> 00:05:19.490 align:start position:0% bolted you have you been out there bones 00:05:19.490 --> 00:05:21.190 align:start position:0% bolted you have you been out there bones are<00:05:19.670> a<00:05:19.790> very<00:05:20.030> important<00:05:20.630> part<00:05:20.810> of<00:05:21.050> the 00:05:21.190 --> 00:05:21.200 align:start position:0% are a very important part of the 00:05:21.200 --> 00:05:23.950 align:start position:0% are a very important part of the democratic<00:05:21.770> process<00:05:22.010> and<00:05:23.000> politicians<00:05:23.720> watch 00:05:23.950 --> 00:05:23.960 align:start position:0% democratic process and politicians watch 00:05:23.960 --> 00:05:27.070 align:start position:0% democratic process and politicians watch them<00:05:24.290> very<00:05:24.320> closely<00:05:25.480> even<00:05:26.480> if<00:05:26.570> they<00:05:26.690> deny<00:05:26.900> it 00:05:27.070 --> 00:05:27.080 align:start position:0% them very closely even if they deny it 00:05:27.080 --> 00:05:30.790 align:start position:0% them very closely even if they deny it in<00:05:27.350> public<00:05:28.690> number<00:05:29.690> one<00:05:29.870> it<00:05:30.110> depends<00:05:30.560> on<00:05:30.650> who 00:05:30.790 --> 00:05:30.800 align:start position:0% in public number one it depends on who 00:05:30.800 --> 00:05:32.560 align:start position:0% in public number one it depends on who they're<00:05:30.980> talking<00:05:31.010> to<00:05:31.490> if<00:05:31.820> they're<00:05:32.000> talking<00:05:32.450> to 00:05:32.560 --> 00:05:32.570 align:start position:0% they're talking to if they're talking to 00:05:32.570 --> 00:05:35.290 align:start position:0% they're talking to if they're talking to the<00:05:32.720> television<00:05:33.170> or<00:05:34.100> to<00:05:34.550> a<00:05:34.580> public<00:05:35.090> audience 00:05:35.290 --> 00:05:35.300 align:start position:0% the television or to a public audience 00:05:35.300 --> 00:05:38.170 align:start position:0% the television or to a public audience they<00:05:36.230> don't<00:05:36.440> believe<00:05:36.620> in<00:05:36.800> opinion<00:05:37.190> polls<00:05:37.220> they 00:05:38.170 --> 00:05:38.180 align:start position:0% they don't believe in opinion polls they 00:05:38.180 --> 00:05:42.580 align:start position:0% they don't believe in opinion polls they believe<00:05:38.480> in<00:05:38.690> that<00:05:39.760> the<00:05:40.760> second<00:05:41.210> is<00:05:41.390> that<00:05:42.200> if 00:05:42.580 --> 00:05:42.590 align:start position:0% believe in that the second is that if 00:05:42.590 --> 00:05:46.180 align:start position:0% believe in that the second is that if they're<00:05:43.280> talking<00:05:43.730> in<00:05:43.880> private<00:05:44.830> they<00:05:45.830> believe 00:05:46.180 --> 00:05:46.190 align:start position:0% they're talking in private they believe 00:05:46.190 --> 00:05:49.900 align:start position:0% they're talking in private they believe them<00:05:46.780> beyond<00:05:47.780> all<00:05:48.140> recognition<00:05:48.350> and<00:05:49.160> the 00:05:49.900 --> 00:05:49.910 align:start position:0% them beyond all recognition and the 00:05:49.910 --> 00:05:53.730 align:start position:0% them beyond all recognition and the third<00:05:50.210> one<00:05:50.450> if<00:05:50.660> they<00:05:51.230> change<00:05:51.500> their<00:05:51.830> policies 00:05:53.730 --> 00:05:53.740 align:start position:0% third one if they change their policies 00:05:53.740 --> 00:05:57.310 align:start position:0% third one if they change their policies I'm<00:05:54.740> delighted<00:05:55.390> because<00:05:56.390> it<00:05:56.810> says<00:05:57.050> they're 00:05:57.310 --> 00:05:57.320 align:start position:0% I'm delighted because it says they're 00:05:57.320 --> 00:05:59.680 align:start position:0% I'm delighted because it says they're listening<00:05:57.560> to<00:05:57.800> public<00:05:58.340> opinion<00:05:58.490> and<00:05:59.090> if 00:05:59.680 --> 00:05:59.690 align:start position:0% listening to public opinion and if 00:05:59.690 --> 00:06:01.570 align:start position:0% listening to public opinion and if they're<00:05:59.900> not<00:06:00.050> there<00:06:00.530> to<00:06:00.830> represent<00:06:01.070> the 00:06:01.570 --> 00:06:01.580 align:start position:0% they're not there to represent the 00:06:01.580 --> 00:06:05.780 align:start position:0% they're not there to represent the public<00:06:02.030> what<00:06:02.270> are<00:06:02.480> they<00:06:02.690> doing<00:06:03.080> in<00:06:03.260> Parliament 00:06:05.780 --> 00:06:05.790 align:start position:0% public what are they doing in Parliament 00:06:05.790 --> 00:06:08.240 align:start position:0% public what are they doing in Parliament one<00:06:06.450> widely<00:06:06.840> used<00:06:07.050> scheme<00:06:07.350> including<00:06:08.130> in 00:06:08.240 --> 00:06:08.250 align:start position:0% one widely used scheme including in 00:06:08.250 --> 00:06:10.760 align:start position:0% one widely used scheme including in political<00:06:08.730> polling<00:06:09.000> is<00:06:09.480> stratified<00:06:10.080> random 00:06:10.760 --> 00:06:10.770 align:start position:0% political polling is stratified random 00:06:10.770 --> 00:06:13.160 align:start position:0% political polling is stratified random sampling<00:06:11.190> where<00:06:11.910> the<00:06:12.030> population<00:06:12.630> is<00:06:12.720> divided 00:06:13.160 --> 00:06:13.170 align:start position:0% sampling where the population is divided 00:06:13.170 --> 00:06:15.550 align:start position:0% sampling where the population is divided into<00:06:13.500> a<00:06:13.620> number<00:06:13.710> of<00:06:14.190> non<00:06:14.760> overlapping 00:06:15.550 --> 00:06:15.560 align:start position:0% into a number of non overlapping 00:06:15.560 --> 00:06:18.260 align:start position:0% into a number of non overlapping subpopulations<00:06:16.560> called<00:06:16.860> structure<00:06:17.520> and<00:06:17.730> then 00:06:18.260 --> 00:06:18.270 align:start position:0% subpopulations called structure and then 00:06:18.270 --> 00:06:20.330 align:start position:0% subpopulations called structure and then a<00:06:18.360> random<00:06:18.840> sample<00:06:19.110> is<00:06:19.500> taken<00:06:19.530> out<00:06:20.040> of<00:06:20.190> each 00:06:20.330 --> 00:06:20.340 align:start position:0% a random sample is taken out of each 00:06:20.340 --> 00:06:23.240 align:start position:0% a random sample is taken out of each stratum<00:06:21.050> the<00:06:22.050> basis<00:06:22.470> of<00:06:22.500> the<00:06:22.800> stratification 00:06:23.240 --> 00:06:23.250 align:start position:0% stratum the basis of the stratification 00:06:23.250 --> 00:06:28.880 align:start position:0% stratum the basis of the stratification might<00:06:23.910> be<00:06:24.090> age<00:06:24.830> income<00:06:25.940> residents<00:06:27.890> this 00:06:28.880 --> 00:06:28.890 align:start position:0% might be age income residents this 00:06:28.890 --> 00:06:31.010 align:start position:0% might be age income residents this increases<00:06:29.400> the<00:06:29.400> reliability<00:06:29.670> of<00:06:30.330> the<00:06:30.540> sample 00:06:31.010 --> 00:06:31.020 align:start position:0% increases the reliability of the sample 00:06:31.020 --> 00:06:32.930 align:start position:0% increases the reliability of the sample making<00:06:31.350> it<00:06:31.470> more<00:06:31.770> representative<00:06:32.670> of<00:06:32.760> the 00:06:32.930 --> 00:06:32.940 align:start position:0% making it more representative of the 00:06:32.940 --> 00:06:35.950 align:start position:0% making it more representative of the population<00:06:33.290> while<00:06:34.290> saving<00:06:34.590> some<00:06:34.920> costs 00:06:35.950 --> 00:06:35.960 align:start position:0% population while saving some costs 00:06:35.960 --> 00:06:38.480 align:start position:0% population while saving some costs samples<00:06:36.960> from<00:06:37.080> different<00:06:37.140> strata<00:06:37.740> can<00:06:38.370> be 00:06:38.480 --> 00:06:38.490 align:start position:0% samples from different strata can be 00:06:38.490 --> 00:06:41.120 align:start position:0% samples from different strata can be different<00:06:38.790> in<00:06:38.970> size<00:06:39.680> non-proportional<00:06:40.680> if 00:06:41.120 --> 00:06:41.130 align:start position:0% different in size non-proportional if 00:06:41.130 --> 00:06:42.770 align:start position:0% different in size non-proportional if they<00:06:41.340> need<00:06:41.550> to<00:06:41.580> reflect<00:06:41.970> the<00:06:42.210> different<00:06:42.390> size 00:06:42.770 --> 00:06:42.780 align:start position:0% they need to reflect the different size 00:06:42.780 --> 00:06:45.320 align:start position:0% they need to reflect the different size or<00:06:43.170> significance<00:06:43.890> of<00:06:44.010> the<00:06:44.160> strata<00:06:44.460> for 00:06:45.320 --> 00:06:45.330 align:start position:0% or significance of the strata for 00:06:45.330 --> 00:06:46.970 align:start position:0% or significance of the strata for example<00:06:45.780> if<00:06:45.960> the<00:06:46.080> proportion<00:06:46.710> of<00:06:46.800> population 00:06:46.970 --> 00:06:46.980 align:start position:0% example if the proportion of population 00:06:46.980 --> 00:06:50.360 align:start position:0% example if the proportion of population over<00:06:47.730> 60<00:06:48.210> is<00:06:48.330> larger<00:06:48.960> or<00:06:49.110> if<00:06:49.740> a<00:06:49.860> larger 00:06:50.360 --> 00:06:50.370 align:start position:0% over 60 is larger or if a larger 00:06:50.370 --> 00:06:52.730 align:start position:0% over 60 is larger or if a larger proportion<00:06:50.880> of<00:06:51.000> this<00:06:51.180> age<00:06:51.420> votes<00:06:51.900> the 00:06:52.730 --> 00:06:52.740 align:start position:0% proportion of this age votes the 00:06:52.740 --> 00:06:54.500 align:start position:0% proportion of this age votes the pollsters<00:06:53.190> might<00:06:53.430> then<00:06:53.700> want<00:06:53.730> to<00:06:54.000> interview 00:06:54.500 --> 00:06:54.510 align:start position:0% pollsters might then want to interview 00:06:54.510 --> 00:06:58.250 align:start position:0% pollsters might then want to interview more<00:06:54.840> of<00:06:55.020> those<00:06:55.260> and<00:06:55.560> less<00:06:56.310> of<00:06:56.520> Sadie<00:06:56.850> 18<00:06:57.570> to<00:06:57.750> 25 00:06:58.250 --> 00:06:58.260 align:start position:0% more of those and less of Sadie 18 to 25 00:06:58.260 --> 00:07:01.190 align:start position:0% more of those and less of Sadie 18 to 25 group<00:06:59.210> another<00:07:00.210> form<00:07:00.510> of<00:07:00.630> stratified 00:07:01.190 --> 00:07:01.200 align:start position:0% group another form of stratified 00:07:01.200 --> 00:07:03.800 align:start position:0% group another form of stratified sampling<00:07:01.680> is<00:07:01.920> quota<00:07:02.340> sampling<00:07:02.820> where 00:07:03.800 --> 00:07:03.810 align:start position:0% sampling is quota sampling where 00:07:03.810 --> 00:07:05.900 align:start position:0% sampling is quota sampling where personal<00:07:04.410> judgment<00:07:04.890> plays<00:07:05.130> a<00:07:05.160> significant 00:07:05.900 --> 00:07:05.910 align:start position:0% personal judgment plays a significant 00:07:05.910 --> 00:07:08.690 align:start position:0% personal judgment plays a significant role<00:07:05.970> in<00:07:06.180> the<00:07:06.390> selection<00:07:06.990> process<00:07:07.700> such 00:07:08.690 --> 00:07:08.700 align:start position:0% role in the selection process such 00:07:08.700 --> 00:07:10.760 align:start position:0% role in the selection process such samples<00:07:08.940> we<00:07:09.300> call<00:07:09.510> judgment<00:07:10.170> or<00:07:10.380> non 00:07:10.760 --> 00:07:10.770 align:start position:0% samples we call judgment or non 00:07:10.770 --> 00:07:14.170 align:start position:0% samples we call judgment or non probability<00:07:11.550> samples<00:07:12.530> if<00:07:13.530> we<00:07:13.800> use 00:07:14.170 --> 00:07:14.180 align:start position:0% probability samples if we use 00:07:14.180 --> 00:07:17.510 align:start position:0% probability samples if we use non-probability<00:07:15.180> sampling<00:07:16.140> we<00:07:16.920> can't<00:07:17.250> use 00:07:17.510 --> 00:07:17.520 align:start position:0% non-probability sampling we can't use 00:07:17.520 --> 00:07:20.390 align:start position:0% non-probability sampling we can't use quite<00:07:17.910> the<00:07:18.150> same<00:07:18.390> formula<00:07:19.050> that<00:07:19.500> we<00:07:19.710> use<00:07:19.950> if<00:07:20.250> we 00:07:20.390 --> 00:07:20.400 align:start position:0% quite the same formula that we use if we 00:07:20.400 --> 00:07:24.290 align:start position:0% quite the same formula that we use if we have<00:07:20.750> probability<00:07:21.750> sampling 00:07:24.290 --> 00:07:24.300 align:start position:0% 00:07:24.300 --> 00:07:26.689 align:start position:0% do<00:07:25.169> hereby<00:07:25.409> you<00:07:25.560> notice<00:07:26.099> that<00:07:26.280> the<00:07:26.370> number<00:07:26.669> of 00:07:26.689 --> 00:07:26.699 align:start position:0% do hereby you notice that the number of 00:07:26.699 --> 00:07:28.790 align:start position:0% do hereby you notice that the number of votes<00:07:26.940> recorded<00:07:27.599> for<00:07:27.810> each<00:07:27.960> candidate<00:07:28.620> at<00:07:28.770> the 00:07:28.790 --> 00:07:28.800 align:start position:0% votes recorded for each candidate at the 00:07:28.800 --> 00:07:31.369 align:start position:0% votes recorded for each candidate at the same<00:07:29.129> election<00:07:29.550> is<00:07:29.909> following<00:07:30.659> Caroline<00:07:31.080> won 00:07:31.369 --> 00:07:31.379 align:start position:0% same election is following Caroline won 00:07:31.379 --> 00:07:34.180 align:start position:0% same election is following Caroline won the<00:07:31.530> election<00:07:32.039> let's<00:07:32.970> look<00:07:33.120> at<00:07:33.270> what<00:07:33.419> happened 00:07:34.180 --> 00:07:34.190 align:start position:0% the election let's look at what happened 00:07:34.190 --> 00:07:37.550 align:start position:0% the election let's look at what happened we<00:07:35.190> have<00:07:35.340> two<00:07:35.580> tables<00:07:36.000> of<00:07:36.120> data<00:07:36.150> here<00:07:36.560> the 00:07:37.550 --> 00:07:37.560 align:start position:0% we have two tables of data here the 00:07:37.560 --> 00:07:39.980 align:start position:0% we have two tables of data here the first<00:07:37.860> is<00:07:38.159> a<00:07:38.520> national<00:07:39.030> opinion<00:07:39.300> poll<00:07:39.690> taken 00:07:39.980 --> 00:07:39.990 align:start position:0% first is a national opinion poll taken 00:07:39.990 --> 00:07:42.800 align:start position:0% first is a national opinion poll taken very<00:07:40.319> close<00:07:40.620> to<00:07:40.860> the<00:07:40.949> local<00:07:41.280> elections<00:07:41.810> the 00:07:42.800 --> 00:07:42.810 align:start position:0% very close to the local elections the 00:07:42.810 --> 00:07:45.260 align:start position:0% very close to the local elections the second<00:07:43.259> is<00:07:43.440> the<00:07:43.800> result<00:07:44.370> in<00:07:44.520> Brighton<00:07:44.940> hill<00:07:45.120> in 00:07:45.260 --> 00:07:45.270 align:start position:0% second is the result in Brighton hill in 00:07:45.270 --> 00:07:48.950 align:start position:0% second is the result in Brighton hill in basingstoke<00:07:45.840> as<00:07:46.460> you<00:07:47.460> can<00:07:47.610> see<00:07:47.819> the<00:07:48.569> results 00:07:48.950 --> 00:07:48.960 align:start position:0% basingstoke as you can see the results 00:07:48.960 --> 00:07:52.070 align:start position:0% basingstoke as you can see the results are<00:07:49.080> somewhat<00:07:49.379> different<00:07:50.750> what<00:07:51.750> are<00:07:51.840> the<00:07:51.900> main 00:07:52.070 --> 00:07:52.080 align:start position:0% are somewhat different what are the main 00:07:52.080 --> 00:07:55.490 align:start position:0% are somewhat different what are the main reasons<00:07:53.270> first<00:07:54.270> Brighton<00:07:54.810> Hill<00:07:55.020> is<00:07:55.229> not 00:07:55.490 --> 00:07:55.500 align:start position:0% reasons first Brighton Hill is not 00:07:55.500 --> 00:07:58.300 align:start position:0% reasons first Brighton Hill is not representative<00:07:56.190> of<00:07:56.430> the<00:07:56.580> National<00:07:56.969> picture 00:07:58.300 --> 00:07:58.310 align:start position:0% representative of the National picture 00:07:58.310 --> 00:08:00.499 align:start position:0% representative of the National picture second<00:07:59.310> people<00:07:59.789> may<00:07:59.909> have<00:07:59.940> changed<00:08:00.300> their 00:08:00.499 --> 00:08:00.509 align:start position:0% second people may have changed their 00:08:00.509 --> 00:08:03.459 align:start position:0% second people may have changed their minds<00:08:00.930> between<00:08:01.110> being<00:08:01.560> polled<00:08:01.889> and<00:08:02.219> voting 00:08:03.459 --> 00:08:03.469 align:start position:0% minds between being polled and voting 00:08:03.469 --> 00:08:06.260 align:start position:0% minds between being polled and voting third<00:08:04.469> some<00:08:05.310> of<00:08:05.460> those<00:08:05.580> who<00:08:05.819> said<00:08:06.090> they'd 00:08:06.260 --> 00:08:06.270 align:start position:0% third some of those who said they'd 00:08:06.270 --> 00:08:08.809 align:start position:0% third some of those who said they'd support<00:08:06.629> a<00:08:06.780> particular<00:08:06.960> party<00:08:07.740> may<00:08:08.520> not<00:08:08.699> have 00:08:08.809 --> 00:08:08.819 align:start position:0% support a particular party may not have 00:08:08.819 --> 00:08:10.700 align:start position:0% support a particular party may not have wanted<00:08:08.969> to<00:08:09.270> support<00:08:09.659> a<00:08:09.780> particular<00:08:09.930> local 00:08:10.700 --> 00:08:10.710 align:start position:0% wanted to support a particular local 00:08:10.710 --> 00:08:14.749 align:start position:0% wanted to support a particular local candidate<00:08:11.190> or<00:08:11.370> vice<00:08:11.729> versa<00:08:12.979> finally<00:08:13.979> people 00:08:14.749 --> 00:08:14.759 align:start position:0% candidate or vice versa finally people 00:08:14.759 --> 00:08:16.279 align:start position:0% candidate or vice versa finally people may<00:08:14.969> not<00:08:15.240> have<00:08:15.389> bothered<00:08:15.569> to<00:08:15.840> vote<00:08:15.960> at<00:08:16.169> all 00:08:16.279 --> 00:08:16.289 align:start position:0% may not have bothered to vote at all 00:08:16.289 --> 00:08:19.730 align:start position:0% may not have bothered to vote at all even<00:08:17.039> though<00:08:17.159> they<00:08:17.310> said<00:08:17.550> they<00:08:17.699> would<00:08:18.620> but<00:08:19.620> are 00:08:19.730 --> 00:08:19.740 align:start position:0% even though they said they would but are 00:08:19.740 --> 00:08:22.629 align:start position:0% even though they said they would but are the<00:08:19.889> opinion<00:08:20.340> polls<00:08:20.520> themselves<00:08:21.180> reliable 00:08:22.629 --> 00:08:22.639 align:start position:0% the opinion polls themselves reliable 00:08:22.639 --> 00:08:25.070 align:start position:0% the opinion polls themselves reliable the<00:08:23.639> polls<00:08:23.849> conducted<00:08:24.120> by<00:08:24.539> the<00:08:24.569> local 00:08:25.070 --> 00:08:25.080 align:start position:0% the polls conducted by the local 00:08:25.080 --> 00:08:27.649 align:start position:0% the polls conducted by the local politicians<00:08:25.830> I've<00:08:26.520> got<00:08:26.729> an<00:08:26.849> element<00:08:27.240> of<00:08:27.360> bias 00:08:27.649 --> 00:08:27.659 align:start position:0% politicians I've got an element of bias 00:08:27.659 --> 00:08:30.200 align:start position:0% politicians I've got an element of bias in<00:08:27.960> it<00:08:28.080> because<00:08:28.289> people<00:08:29.279> are<00:08:29.370> inclined<00:08:29.789> to<00:08:29.969> say 00:08:30.200 --> 00:08:30.210 align:start position:0% in it because people are inclined to say 00:08:30.210 --> 00:08:32.120 align:start position:0% in it because people are inclined to say whatever<00:08:30.509> the<00:08:30.840> local<00:08:31.199> interviewer<00:08:31.620> wants<00:08:31.889> to 00:08:32.120 --> 00:08:32.130 align:start position:0% whatever the local interviewer wants to 00:08:32.130 --> 00:08:36.589 align:start position:0% whatever the local interviewer wants to hear<00:08:32.550> do<00:08:33.390> you<00:08:33.479> worry<00:08:33.719> that<00:08:35.089> people<00:08:36.089> will<00:08:36.360> say 00:08:36.589 --> 00:08:36.599 align:start position:0% hear do you worry that people will say 00:08:36.599 --> 00:08:39.230 align:start position:0% hear do you worry that people will say anything<00:08:37.140> to<00:08:37.860> please<00:08:38.099> the<00:08:38.490> interviewer<00:08:39.029> oh 00:08:39.230 --> 00:08:39.240 align:start position:0% anything to please the interviewer oh 00:08:39.240 --> 00:08:43.490 align:start position:0% anything to please the interviewer oh yes<00:08:39.690> that<00:08:39.870> does<00:08:40.020> happen<00:08:41.329> yes<00:08:42.329> I<00:08:42.719> mean<00:08:43.110> by<00:08:43.440> the 00:08:43.490 --> 00:08:43.500 align:start position:0% yes that does happen yes I mean by the 00:08:43.500 --> 00:08:45.139 align:start position:0% yes that does happen yes I mean by the tall<00:08:43.770> we<00:08:43.860> put<00:08:44.070> person<00:08:44.700> knocks<00:08:44.880> on<00:08:45.029> the<00:08:45.120> door 00:08:45.139 --> 00:08:45.149 align:start position:0% tall we put person knocks on the door 00:08:45.149 --> 00:08:46.790 align:start position:0% tall we put person knocks on the door they'll<00:08:45.480> say<00:08:45.630> they<00:08:45.779> can't<00:08:45.930> devote<00:08:46.140> to<00:08:46.350> even 00:08:46.790 --> 00:08:46.800 align:start position:0% they'll say they can't devote to even 00:08:46.800 --> 00:08:47.930 align:start position:0% they'll say they can't devote to even the<00:08:46.890> Labour<00:08:47.100> person<00:08:47.550> knocks<00:08:47.700> on<00:08:47.820> the<00:08:47.910> door 00:08:47.930 --> 00:08:47.940 align:start position:0% the Labour person knocks on the door 00:08:47.940 --> 00:08:50.180 align:start position:0% the Labour person knocks on the door they<00:08:48.209> get<00:08:48.360> so<00:08:48.480> again<00:08:48.720> etc<00:08:48.810> but<00:08:49.770> yes<00:08:49.980> at<00:08:50.130> the 00:08:50.180 --> 00:08:50.190 align:start position:0% they get so again etc but yes at the 00:08:50.190 --> 00:08:51.500 align:start position:0% they get so again etc but yes at the same<00:08:50.399> time<00:08:50.579> they<00:08:50.790> know<00:08:50.940> at<00:08:51.089> the<00:08:51.180> end<00:08:51.209> of<00:08:51.420> the 00:08:51.500 --> 00:08:51.510 align:start position:0% same time they know at the end of the 00:08:51.510 --> 00:08:53.090 align:start position:0% same time they know at the end of the day<00:08:51.630> people<00:08:51.930> might<00:08:52.200> be<00:08:52.350> telling<00:08:52.529> the<00:08:52.680> rubbish 00:08:53.090 --> 00:08:53.100 align:start position:0% day people might be telling the rubbish 00:08:53.100 --> 00:08:55.760 align:start position:0% day people might be telling the rubbish so<00:08:53.730> you<00:08:53.820> know<00:08:53.970> there's<00:08:54.240> an<00:08:54.420> element<00:08:55.380> about 00:08:55.760 --> 00:08:55.770 align:start position:0% so you know there's an element about 00:08:55.770 --> 00:09:00.050 align:start position:0% so you know there's an element about reliability<00:08:56.220> obviously<00:08:57.209> yes<00:08:57.390> yes<00:08:58.970> but<00:08:59.970> is 00:09:00.050 --> 00:09:00.060 align:start position:0% reliability obviously yes yes but is 00:09:00.060 --> 00:09:02.480 align:start position:0% reliability obviously yes yes but is there<00:09:00.209> a<00:09:00.270> problem<00:09:00.630> about<00:09:01.170> the<00:09:01.490> professional 00:09:02.480 --> 00:09:02.490 align:start position:0% there a problem about the professional 00:09:02.490 --> 00:09:05.800 align:start position:0% there a problem about the professional pollsters<00:09:03.620> even<00:09:04.620> there<00:09:04.829> they<00:09:05.010> may<00:09:05.310> be<00:09:05.370> an 00:09:05.800 --> 00:09:05.810 align:start position:0% pollsters even there they may be an 00:09:05.810 --> 00:09:08.900 align:start position:0% pollsters even there they may be an interview<00:09:06.810> effect<00:09:07.199> where<00:09:08.160> people<00:09:08.520> listen<00:09:08.880> to 00:09:08.900 --> 00:09:08.910 align:start position:0% interview effect where people listen to 00:09:08.910 --> 00:09:11.620 align:start position:0% interview effect where people listen to the<00:09:09.120> tone<00:09:09.420> of<00:09:09.449> voice<00:09:09.660> of<00:09:10.290> the<00:09:10.620> interviewer<00:09:11.220> and 00:09:11.620 --> 00:09:11.630 align:start position:0% the tone of voice of the interviewer and 00:09:11.630 --> 00:09:15.260 align:start position:0% the tone of voice of the interviewer and respond<00:09:12.630> still<00:09:13.470> to<00:09:14.190> what<00:09:14.370> they<00:09:14.550> think<00:09:15.000> the 00:09:15.260 --> 00:09:15.270 align:start position:0% respond still to what they think the 00:09:15.270 --> 00:09:18.860 align:start position:0% respond still to what they think the interviewer<00:09:15.779> wants<00:09:15.990> to<00:09:16.260> hear<00:09:17.690> so<00:09:18.690> there<00:09:18.839> are 00:09:18.860 --> 00:09:18.870 align:start position:0% interviewer wants to hear so there are 00:09:18.870 --> 00:09:22.610 align:start position:0% interviewer wants to hear so there are two<00:09:19.140> issues<00:09:19.320> here<00:09:19.850> one<00:09:20.850> is<00:09:21.120> the<00:09:21.750> reliability 00:09:22.610 --> 00:09:22.620 align:start position:0% two issues here one is the reliability 00:09:22.620 --> 00:09:24.949 align:start position:0% two issues here one is the reliability of<00:09:22.649> the<00:09:22.890> sample<00:09:23.430> accurately<00:09:24.209> reflecting 00:09:24.949 --> 00:09:24.959 align:start position:0% of the sample accurately reflecting 00:09:24.959 --> 00:09:29.860 align:start position:0% of the sample accurately reflecting opinion<00:09:26.029> to<00:09:27.029> is<00:09:27.959> the<00:09:28.230> value<00:09:28.680> of<00:09:28.709> the<00:09:29.010> polis<00:09:29.610> of 00:09:29.860 --> 00:09:29.870 align:start position:0% opinion to is the value of the polis of 00:09:29.870 --> 00:09:32.480 align:start position:0% opinion to is the value of the polis of forecasting<00:09:30.870> tool<00:09:31.170> good<00:09:31.470> when<00:09:32.250> voting 00:09:32.480 --> 00:09:32.490 align:start position:0% forecasting tool good when voting 00:09:32.490 --> 00:09:35.940 align:start position:0% forecasting tool good when voting intentions<00:09:33.149> can<00:09:33.300> change<00:09:33.630> quickly 00:09:35.940 --> 00:09:35.950 align:start position:0% 00:09:35.950 --> 00:09:39.270 align:start position:0% thinking<00:09:36.580> about<00:09:36.670> that<00:09:36.970> first<00:09:37.300> issue<00:09:38.010> are<00:09:39.010> they 00:09:39.270 --> 00:09:39.280 align:start position:0% thinking about that first issue are they 00:09:39.280 --> 00:09:42.390 align:start position:0% thinking about that first issue are they accurate<00:09:39.960> just<00:09:40.960> how<00:09:41.170> big<00:09:41.470> a<00:09:41.530> sample<00:09:42.190> do<00:09:42.280> you 00:09:42.390 --> 00:09:42.400 align:start position:0% accurate just how big a sample do you 00:09:42.400 --> 00:09:44.760 align:start position:0% accurate just how big a sample do you need<00:09:42.700> to<00:09:42.940> get<00:09:43.090> a<00:09:43.120> useful<00:09:43.690> estimate<00:09:44.170> of<00:09:44.320> the<00:09:44.620> way 00:09:44.760 --> 00:09:44.770 align:start position:0% need to get a useful estimate of the way 00:09:44.770 --> 00:09:47.550 align:start position:0% need to get a useful estimate of the way people<00:09:45.190> are<00:09:45.250> thinking<00:09:45.490> about<00:09:45.730> voting<00:09:46.560> our 00:09:47.550 --> 00:09:47.560 align:start position:0% people are thinking about voting our 00:09:47.560 --> 00:09:50.250 align:start position:0% people are thinking about voting our understanding<00:09:48.460> of<00:09:48.550> Statistics<00:09:49.180> will<00:09:49.960> give<00:09:50.140> us 00:09:50.250 --> 00:09:50.260 align:start position:0% understanding of Statistics will give us 00:09:50.260 --> 00:09:53.880 align:start position:0% understanding of Statistics will give us an<00:09:50.380> insight<00:09:50.740> into<00:09:51.070> this<00:09:52.470> now<00:09:53.470> here's<00:09:53.770> the 00:09:53.880 --> 00:09:53.890 align:start position:0% an insight into this now here's the 00:09:53.890 --> 00:09:55.370 align:start position:0% an insight into this now here's the sampling<00:09:54.280> problem<00:09:54.670> we<00:09:54.760> want<00:09:54.880> to<00:09:55.000> solve 00:09:55.370 --> 00:09:55.380 align:start position:0% sampling problem we want to solve 00:09:55.380 --> 00:09:59.820 align:start position:0% sampling problem we want to solve assuming<00:09:56.380> we've<00:09:56.530> got<00:09:56.740> say<00:09:57.840> 1569<00:09:58.840> electors<00:09:59.590> and 00:09:59.820 --> 00:09:59.830 align:start position:0% assuming we've got say 1569 electors and 00:09:59.830 --> 00:10:01.800 align:start position:0% assuming we've got say 1569 electors and twenty<00:10:00.520> percent<00:10:00.820> are<00:10:01.030> known<00:10:01.180> to<00:10:01.360> be<00:10:01.390> Labour 00:10:01.800 --> 00:10:01.810 align:start position:0% twenty percent are known to be Labour 00:10:01.810 --> 00:10:04.320 align:start position:0% twenty percent are known to be Labour Party<00:10:02.350> supporters<00:10:02.950> how<00:10:03.610> many<00:10:03.790> people<00:10:04.150> would 00:10:04.320 --> 00:10:04.330 align:start position:0% Party supporters how many people would 00:10:04.330 --> 00:10:07.230 align:start position:0% Party supporters how many people would we<00:10:04.510> have<00:10:04.690> to<00:10:04.870> sample<00:10:05.170> to<00:10:05.980> be<00:10:06.010> confident<00:10:06.730> we've 00:10:07.230 --> 00:10:07.240 align:start position:0% we have to sample to be confident we've 00:10:07.240 --> 00:10:08.850 align:start position:0% we have to sample to be confident we've got<00:10:07.420> a<00:10:07.450> suitably<00:10:07.960> representative 00:10:08.850 --> 00:10:08.860 align:start position:0% got a suitably representative 00:10:08.860 --> 00:10:12.100 align:start position:0% got a suitably representative cross-section<00:10:09.100> of<00:10:09.730> voting<00:10:10.120> opinion 00:10:12.100 --> 00:10:12.110 align:start position:0% cross-section of voting opinion 00:10:12.110 --> 00:10:14.290 align:start position:0% cross-section of voting opinion there's<00:10:12.769> a<00:10:12.890> formula<00:10:13.190> that<00:10:13.370> we<00:10:13.640> can<00:10:13.820> use<00:10:13.850> to 00:10:14.290 --> 00:10:14.300 align:start position:0% there's a formula that we can use to 00:10:14.300 --> 00:10:17.280 align:start position:0% there's a formula that we can use to find<00:10:14.750> the<00:10:15.290> number<00:10:15.589> of<00:10:15.740> voters<00:10:15.920> we<00:10:16.220> need<00:10:16.399> to<00:10:16.550> ask 00:10:17.280 --> 00:10:17.290 align:start position:0% find the number of voters we need to ask 00:10:17.290 --> 00:10:22.060 align:start position:0% find the number of voters we need to ask the<00:10:18.290> size<00:10:18.620> of<00:10:18.890> the<00:10:18.980> sample<00:10:19.190> will<00:10:19.579> call<00:10:19.820> n<00:10:20.180> and<00:10:21.070> n 00:10:22.060 --> 00:10:22.070 align:start position:0% the size of the sample will call n and n 00:10:22.070 --> 00:10:26.579 align:start position:0% the size of the sample will call n and n must<00:10:22.519> satisfy<00:10:23.000> the<00:10:23.060> equation<00:10:24.100> Z<00:10:25.100> sub<00:10:25.550> naught 00:10:26.579 --> 00:10:26.589 align:start position:0% must satisfy the equation Z sub naught 00:10:26.589 --> 00:10:31.900 align:start position:0% must satisfy the equation Z sub naught times<00:10:28.089> the<00:10:29.089> square<00:10:29.240> root<00:10:29.510> of<00:10:29.959> P<00:10:30.649> Q<00:10:30.920> divided<00:10:31.700> by 00:10:31.900 --> 00:10:31.910 align:start position:0% times the square root of P Q divided by 00:10:31.910 --> 00:10:37.509 align:start position:0% times the square root of P Q divided by n<00:10:32.560> equals<00:10:33.560> e<00:10:34.329> where<00:10:35.329> p<00:10:35.570> is<00:10:35.779> the<00:10:36.529> proportion 00:10:37.509 --> 00:10:37.519 align:start position:0% n equals e where p is the proportion 00:10:37.519 --> 00:10:40.360 align:start position:0% n equals e where p is the proportion favoring<00:10:38.360> a<00:10:38.450> particular<00:10:38.720> party<00:10:39.470> in<00:10:39.649> our<00:10:40.130> case 00:10:40.360 --> 00:10:40.370 align:start position:0% favoring a particular party in our case 00:10:40.370 --> 00:10:44.470 align:start position:0% favoring a particular party in our case the<00:10:40.399> labor<00:10:40.820> party<00:10:41.649> q<00:10:42.649> is<00:10:42.860> the<00:10:42.890> rest<00:10:43.850> of<00:10:44.149> our 00:10:44.470 --> 00:10:44.480 align:start position:0% the labor party q is the rest of our 00:10:44.480 --> 00:10:46.960 align:start position:0% the labor party q is the rest of our population<00:10:45.050> it<00:10:45.890> might<00:10:46.160> be<00:10:46.339> just<00:10:46.610> the 00:10:46.960 --> 00:10:46.970 align:start position:0% population it might be just the 00:10:46.970 --> 00:10:49.360 align:start position:0% population it might be just the Conservatives<00:10:47.720> or<00:10:47.990> the<00:10:48.620> Conservatives 00:10:49.360 --> 00:10:49.370 align:start position:0% Conservatives or the Conservatives 00:10:49.370 --> 00:10:53.319 align:start position:0% Conservatives or the Conservatives Liberal<00:10:50.060> Democrats<00:10:50.570> and<00:10:50.810> others<00:10:52.089> Z<00:10:53.089> sub 00:10:53.319 --> 00:10:53.329 align:start position:0% Liberal Democrats and others Z sub 00:10:53.329 --> 00:10:57.280 align:start position:0% Liberal Democrats and others Z sub naught<00:10:53.510> is<00:10:54.200> the<00:10:54.230> confidence<00:10:55.070> limit<00:10:55.430> and<00:10:55.930> E<00:10:56.930> is 00:10:57.280 --> 00:10:57.290 align:start position:0% naught is the confidence limit and E is 00:10:57.290 --> 00:11:00.449 align:start position:0% naught is the confidence limit and E is the<00:10:57.459> maximum<00:10:58.459> error<00:10:58.640> that's<00:10:58.940> acceptable 00:11:00.449 --> 00:11:00.459 align:start position:0% the maximum error that's acceptable 00:11:00.459 --> 00:11:04.180 align:start position:0% the maximum error that's acceptable notice<00:11:01.459> that<00:11:01.490> of<00:11:01.820> course<00:11:02.029> p<00:11:02.899> plus<00:11:03.560> q<00:11:03.589> will 00:11:04.180 --> 00:11:04.190 align:start position:0% notice that of course p plus q will 00:11:04.190 --> 00:11:11.560 align:start position:0% notice that of course p plus q will equal<00:11:04.550> one<00:11:04.930> but<00:11:05.930> p<00:11:06.260> times<00:11:06.890> q<00:11:07.160> won't<00:11:10.120> now<00:11:11.120> our 00:11:11.560 --> 00:11:11.570 align:start position:0% equal one but p times q won't now our 00:11:11.570 --> 00:11:15.639 align:start position:0% equal one but p times q won't now our unknown<00:11:12.290> is<00:11:12.500> n<00:11:12.890> so<00:11:13.570> to<00:11:14.570> find<00:11:14.959> then<00:11:15.200> we'll<00:11:15.440> need 00:11:15.639 --> 00:11:15.649 align:start position:0% unknown is n so to find then we'll need 00:11:15.649 --> 00:11:18.400 align:start position:0% unknown is n so to find then we'll need to<00:11:15.860> rearrange<00:11:16.220> our<00:11:16.730> equation<00:11:17.360> with<00:11:17.630> n<00:11:17.930> on<00:11:18.230> the 00:11:18.400 --> 00:11:18.410 align:start position:0% to rearrange our equation with n on the 00:11:18.410 --> 00:11:21.939 align:start position:0% to rearrange our equation with n on the left<00:11:18.649> hand<00:11:18.920> side<00:11:19.990> doing<00:11:20.990> this<00:11:21.170> gives<00:11:21.500> us<00:11:21.740> that 00:11:21.939 --> 00:11:21.949 align:start position:0% left hand side doing this gives us that 00:11:21.949 --> 00:11:29.400 align:start position:0% left hand side doing this gives us that n<00:11:22.660> equals<00:11:24.190> Z<00:11:25.190> sub<00:11:25.459> naught<00:11:25.670> squared<00:11:26.680> times<00:11:27.680> P<00:11:28.220> Q 00:11:29.400 --> 00:11:29.410 align:start position:0% n equals Z sub naught squared times P Q 00:11:29.410 --> 00:11:34.480 align:start position:0% n equals Z sub naught squared times P Q divided<00:11:30.410> by<00:11:30.680> E<00:11:30.980> squared<00:11:32.949> now<00:11:33.949> here<00:11:34.279> we'll 00:11:34.480 --> 00:11:34.490 align:start position:0% divided by E squared now here we'll 00:11:34.490 --> 00:11:41.410 align:start position:0% divided by E squared now here we'll choose<00:11:34.940> that<00:11:35.329> Z<00:11:36.050> sub<00:11:36.350> naught<00:11:36.699> B<00:11:38.350> 1.96<00:11:40.240> we<00:11:41.240> need 00:11:41.410 --> 00:11:41.420 align:start position:0% choose that Z sub naught B 1.96 we need 00:11:41.420 --> 00:11:44.980 align:start position:0% choose that Z sub naught B 1.96 we need to<00:11:41.540> be<00:11:41.690> within<00:11:41.949> 1.96<00:11:42.949> standard<00:11:43.839> deviations<00:11:44.839> of 00:11:44.980 --> 00:11:44.990 align:start position:0% to be within 1.96 standard deviations of 00:11:44.990 --> 00:11:48.880 align:start position:0% to be within 1.96 standard deviations of the<00:11:45.050> mean<00:11:45.850> to<00:11:46.850> give<00:11:47.060> us<00:11:47.649> ninety-five<00:11:48.649> percent 00:11:48.880 --> 00:11:48.890 align:start position:0% the mean to give us ninety-five percent 00:11:48.890 --> 00:11:54.449 align:start position:0% the mean to give us ninety-five percent confidence<00:11:49.790> in<00:11:50.000> our<00:11:50.120> answer<00:11:51.910> p<00:11:52.910> is<00:11:53.060> not<00:11:53.269> point2 00:11:54.449 --> 00:11:54.459 align:start position:0% confidence in our answer p is not point2 00:11:54.459 --> 00:11:58.700 align:start position:0% confidence in our answer p is not point2 so<00:11:55.459> q<00:11:55.790> is<00:11:55.820> nor<00:11:56.180> point<00:11:56.510> eight 00:11:58.700 --> 00:11:58.710 align:start position:0% so q is nor point eight 00:11:58.710 --> 00:12:02.980 align:start position:0% so q is nor point eight and<00:11:58.800> we'll<00:11:59.640> choose<00:12:00.000> to<00:12:00.720> be<00:12:00.900> naught<00:12:01.440> point<00:12:01.800> 01 00:12:02.980 --> 00:12:02.990 align:start position:0% and we'll choose to be naught point 01 00:12:02.990 --> 00:12:07.150 align:start position:0% and we'll choose to be naught point 01 because<00:12:03.990> we<00:12:04.230> want<00:12:04.380> to<00:12:04.560> be<00:12:04.710> within<00:12:05.040> one<00:12:05.640> percent 00:12:07.150 --> 00:12:07.160 align:start position:0% because we want to be within one percent 00:12:07.160 --> 00:12:12.780 align:start position:0% because we want to be within one percent with<00:12:08.160> a<00:12:08.190> ninety-five<00:12:09.150> percent<00:12:09.360> certainty 00:12:12.780 --> 00:12:12.790 align:start position:0% 00:12:12.790 --> 00:12:17.370 align:start position:0% so<00:12:13.480> using<00:12:13.810> these<00:12:13.930> values<00:12:13.990> n<00:12:14.680> equals<00:12:16.380> 1.96 00:12:17.370 --> 00:12:17.380 align:start position:0% so using these values n equals 1.96 00:12:17.380 --> 00:12:24.020 align:start position:0% so using these values n equals 1.96 squared<00:12:18.000> x<00:12:19.000> naught<00:12:19.960> point<00:12:20.170> 16<00:12:21.900> /<00:12:22.900> naught<00:12:23.710> point 00:12:24.020 --> 00:12:24.030 align:start position:0% squared x naught point 16 / naught point 00:12:24.030 --> 00:12:28.590 align:start position:0% squared x naught point 16 / naught point naught<00:12:25.030> naught<00:12:25.300> naught<00:12:25.630> 1<00:12:27.030> which<00:12:28.030> gives<00:12:28.060> us 00:12:28.590 --> 00:12:28.600 align:start position:0% naught naught naught 1 which gives us 00:12:28.600 --> 00:12:34.680 align:start position:0% naught naught naught 1 which gives us not<00:12:29.290> point<00:12:29.620> 06<00:12:29.880> 14<00:12:30.880> 6<00:12:31.120> 56<00:12:32.130> /<00:12:33.420> naught<00:12:34.420> point 00:12:34.680 --> 00:12:34.690 align:start position:0% not point 06 14 6 56 / naught point 00:12:34.690 --> 00:12:39.480 align:start position:0% not point 06 14 6 56 / naught point naught<00:12:35.200> naught<00:12:35.410> naught<00:12:35.710> 1<00:12:36.450> which<00:12:37.450> equals<00:12:38.490> 614 00:12:39.480 --> 00:12:39.490 align:start position:0% naught naught naught 1 which equals 614 00:12:39.490 --> 00:12:43.740 align:start position:0% naught naught naught 1 which equals 614 point<00:12:39.880> 65<00:12:41.220> now<00:12:42.220> we<00:12:42.280> can't<00:12:42.580> interview<00:12:43.060> point<00:12:43.420> 65 00:12:43.740 --> 00:12:43.750 align:start position:0% point 65 now we can't interview point 65 00:12:43.750 --> 00:12:47.220 align:start position:0% point 65 now we can't interview point 65 of<00:12:44.110> a<00:12:44.200> person<00:12:44.650> so<00:12:45.550> the<00:12:46.180> minimum<00:12:46.510> we<00:12:46.810> need<00:12:46.840> to 00:12:47.220 --> 00:12:47.230 align:start position:0% of a person so the minimum we need to 00:12:47.230 --> 00:12:53.040 align:start position:0% of a person so the minimum we need to interview<00:12:47.680> is<00:12:49.200> 615<00:12:50.200> people<00:12:51.000> that's<00:12:52.000> the<00:12:52.750> size 00:12:53.040 --> 00:12:53.050 align:start position:0% interview is 615 people that's the size 00:12:53.050 --> 00:12:57.300 align:start position:0% interview is 615 people that's the size of<00:12:53.290> the<00:12:53.410> sample<00:12:53.680> will<00:12:54.160> need<00:12:56.160> and<00:12:57.160> what 00:12:57.300 --> 00:12:57.310 align:start position:0% of the sample will need and what 00:12:57.310 --> 00:12:59.580 align:start position:0% of the sample will need and what difference<00:12:57.670> would<00:12:57.820> it<00:12:57.880> have<00:12:57.970> made<00:12:58.200> if<00:12:59.200> the 00:12:59.580 --> 00:12:59.590 align:start position:0% difference would it have made if the 00:12:59.590 --> 00:13:02.670 align:start position:0% difference would it have made if the electorate<00:13:00.070> had<00:13:00.220> been<00:13:00.400> say<00:13:01.140> 5,000<00:13:02.140> people<00:13:02.350> and 00:13:02.670 --> 00:13:02.680 align:start position:0% electorate had been say 5,000 people and 00:13:02.680 --> 00:13:06.150 align:start position:0% electorate had been say 5,000 people and the<00:13:03.580> answer<00:13:03.970> perhaps<00:13:04.120> surprisingly<00:13:04.900> is<00:13:05.170> no 00:13:06.150 --> 00:13:06.160 align:start position:0% the answer perhaps surprisingly is no 00:13:06.160 --> 00:13:10.080 align:start position:0% the answer perhaps surprisingly is no difference<00:13:06.550> at<00:13:06.670> all<00:13:07.500> for<00:13:08.500> large<00:13:08.710> numbers<00:13:09.250> the 00:13:10.080 --> 00:13:10.090 align:start position:0% difference at all for large numbers the 00:13:10.090 --> 00:13:13.320 align:start position:0% difference at all for large numbers the sample<00:13:10.570> size<00:13:10.780> is<00:13:11.170> independent<00:13:12.160> of<00:13:12.340> the<00:13:13.090> size 00:13:13.320 --> 00:13:13.330 align:start position:0% sample size is independent of the size 00:13:13.330 --> 00:13:16.280 align:start position:0% sample size is independent of the size of<00:13:13.510> the<00:13:13.600> whole<00:13:13.780> population 00:13:16.280 --> 00:13:16.290 align:start position:0% 00:13:16.290 --> 00:13:18.499 align:start position:0% what<00:13:16.980> difference<00:13:17.399> would<00:13:17.639> it<00:13:17.699> have<00:13:17.790> made<00:13:17.910> if 00:13:18.499 --> 00:13:18.509 align:start position:0% what difference would it have made if 00:13:18.509 --> 00:13:21.499 align:start position:0% what difference would it have made if say<00:13:19.350> thirty<00:13:20.070> percent<00:13:20.490> rather<00:13:20.880> than<00:13:21.120> twenty 00:13:21.499 --> 00:13:21.509 align:start position:0% say thirty percent rather than twenty 00:13:21.509 --> 00:13:24.249 align:start position:0% say thirty percent rather than twenty percent<00:13:21.839> had<00:13:22.350> been<00:13:22.620> Labour<00:13:22.889> Party<00:13:23.339> supporters 00:13:24.249 --> 00:13:24.259 align:start position:0% percent had been Labour Party supporters 00:13:24.259 --> 00:13:27.350 align:start position:0% percent had been Labour Party supporters the<00:13:25.259> answer<00:13:25.589> is<00:13:25.620> quite<00:13:26.040> a<00:13:26.070> bit<00:13:26.399> you<00:13:27.180> might 00:13:27.350 --> 00:13:27.360 align:start position:0% the answer is quite a bit you might 00:13:27.360 --> 00:13:29.870 align:start position:0% the answer is quite a bit you might rework<00:13:27.899> the<00:13:28.079> formula<00:13:28.560> to<00:13:29.100> discover<00:13:29.519> the 00:13:29.870 --> 00:13:29.880 align:start position:0% rework the formula to discover the 00:13:29.880 --> 00:13:32.390 align:start position:0% rework the formula to discover the necessary<00:13:30.420> number<00:13:30.630> now 00:13:32.390 --> 00:13:32.400 align:start position:0% necessary number now 00:13:32.400 --> 00:13:36.140 align:start position:0% necessary number now wastewater<00:13:33.400> tree<00:13:34.380> climate<00:13:35.380> change<00:13:35.410> research 00:13:36.140 --> 00:13:36.150 align:start position:0% wastewater tree climate change research 00:13:36.150 --> 00:13:39.930 align:start position:0% wastewater tree climate change research and<00:13:37.150> elections<00:13:37.750> seem<00:13:38.470> unrelated<00:13:38.920> but<00:13:39.730> they<00:13:39.880> do 00:13:39.930 --> 00:13:39.940 align:start position:0% and elections seem unrelated but they do 00:13:39.940 --> 00:13:43.170 align:start position:0% and elections seem unrelated but they do have<00:13:40.240> statistical<00:13:40.960> similarities<00:13:42.060> often<00:13:43.060> we 00:13:43.170 --> 00:13:43.180 align:start position:0% have statistical similarities often we 00:13:43.180 --> 00:13:45.600 align:start position:0% have statistical similarities often we rely<00:13:43.540> on<00:13:43.690> statistics<00:13:44.350> to<00:13:44.890> make<00:13:45.070> inferences 00:13:45.600 --> 00:13:45.610 align:start position:0% rely on statistics to make inferences 00:13:45.610 --> 00:13:48.240 align:start position:0% rely on statistics to make inferences about<00:13:45.640> larger<00:13:46.420> systems<00:13:46.810> we<00:13:47.680> can't<00:13:47.890> know<00:13:48.100> what 00:13:48.240 --> 00:13:48.250 align:start position:0% about larger systems we can't know what 00:13:48.250 --> 00:13:51.210 align:start position:0% about larger systems we can't know what each<00:13:48.430> voter<00:13:48.730> is<00:13:48.970> thinking<00:13:49.620> we<00:13:50.620> can't<00:13:50.860> judge 00:13:51.210 --> 00:13:51.220 align:start position:0% each voter is thinking we can't judge 00:13:51.220 --> 00:13:53.360 align:start position:0% each voter is thinking we can't judge climate<00:13:51.670> change<00:13:51.910> by<00:13:52.090> looking<00:13:52.390> around<00:13:52.690> forever 00:13:53.360 --> 00:13:53.370 align:start position:0% climate change by looking around forever 00:13:53.370 --> 00:13:55.800 align:start position:0% climate change by looking around forever we<00:13:54.370> can't<00:13:54.610> judge<00:13:54.880> the<00:13:55.090> effectiveness<00:13:55.330> of 00:13:55.800 --> 00:13:55.810 align:start position:0% we can't judge the effectiveness of 00:13:55.810 --> 00:13:57.810 align:start position:0% we can't judge the effectiveness of wastewater<00:13:56.500> treatment<00:13:56.950> with<00:13:57.490> just<00:13:57.670> one 00:13:57.810 --> 00:13:57.820 align:start position:0% wastewater treatment with just one 00:13:57.820 --> 00:14:01.110 align:start position:0% wastewater treatment with just one sample<00:13:58.440> we<00:13:59.440> need<00:13:59.710> to<00:13:59.920> understand<00:14:00.640> the<00:14:00.820> use<00:14:01.090> of 00:14:01.110 --> 00:14:01.120 align:start position:0% sample we need to understand the use of 00:14:01.120 --> 00:14:04.230 align:start position:0% sample we need to understand the use of statistical<00:14:01.620> techniques<00:14:02.620> if<00:14:02.980> we<00:14:03.970> are<00:14:04.060> to 00:14:04.230 --> 00:14:04.240 align:start position:0% statistical techniques if we are to 00:14:04.240 --> 00:14:06.660 align:start position:0% statistical techniques if we are to arrive<00:14:04.390> at<00:14:04.840> the<00:14:04.930> meaningful<00:14:05.560> answers<00:14:06.130> within 00:14:06.660 --> 00:14:06.670 align:start position:0% arrive at the meaningful answers within 00:14:06.670 --> 00:14:20.750 align:start position:0% arrive at the meaningful answers within some<00:14:07.090> known<00:14:07.390> lunch<00:14:07.720> in<00:14:07.990> a<00:14:08.050> very 00:14:20.750 --> 00:14:20.760 align:start position:0% 00:14:20.760 --> 00:14:22.790 align:start position:0% Oh