Chapter 1 Video 2

0:04 okay now we learn that we have a two types of variable one is the qualitative which one another name was categorical variable and the other one is the quantitative the quantitative variable has again two parts one is the discrete variable the other one is the continuous variable okay one is the discrete the other one is the continuous variable those definitions are right here okay um first one is the discrete okay the discrete variable what is the discrete variable is the data can be only taken as particular values like integers discrete variable data values are usually the things you can count okay this is important uh discrete

0:52 variables are the things you can count okay a discrete variable is a quantitative variable either has a finite number of possible values or counter countable number of values that means you can count okay and then the continuous variable that's the next one the continuous variable is the variable the data can be any value in the interval continuous variable has usually the things you measure okay this is the intro you need to remember the continuous data values are the things you measure okay uh the continuous data values are infinite values right that means you cannot count them but they have a

1:37 unlimited possibilities okay now we are going to be talking about the examples okay most variables are pretty clear but some can be a bit tricky for example the tricky one is the time say for example we were looking for how long we are waiting for a worse bus right we can count the minutes and seconds but really those time units are only rounded okay the actually milliseconds nanoseconds extra has an infinite finite number of possibilities in the middle so actually any variable that can be is a time is a continuous right because there is a

2:23 lot of possibilities okay now we are going to be talking about this an example they are saying example talk about which one is the discrete variable which one is the continues okay again discrete and continues is the part of the quantitative quantitative variable we know their answers are going to be numbers then out of those numbers one are going to be discrete the other one are going to be continuous now we are going to try to figure out which one are discrete which one are continuous okay first um i have iq score a ct score height distance uh commuting distance committing shoe size okay the things you can count are

3:09 discrete the things you can measure are going to be the continues right the the the continuous means there have infinite infinite possibilities okay for first one we have iq score it's going to be discrete because iq score is always like 100 110 120 180 like that okay adc atc score right uh that's going to be discrete because it's the same reasoning like um iq score right there is going to be score given height height you measure okay uh it is going to be continuous even though height is five point eight five eight five feet eight inches it's

3:56 really it can be five point six eight two like feed like it's keep going it's not exact it is impossible to measure the exact length okay and distance commuting again it's going to be continuous just like height you cannot say oh i it took me exact five minutes it can be five point one two three like that many seconds or something like that okay shoe size okay this is the place a lot of people get confused a lot of people think shoe size is continues actually the shoe side is discreet why because in this one what is the shoe sizes are eight eight and a half nine nine and a half ten ten and a half there is no shoe size

4:43 which one is going to be 10.25678 like that right we have only certain shoe sizes that's why it is going to be a discrete not continuous does that make sense guys okay um now if you guys look at the next one um this is an example we are going to be doing together here is the weight of a cat what do you guys think unmute yourself and tell me what do you guys consider continues because we are measuring that right yes it is going to be continuous since it is a it is something you measure

5:39 okay number of fleas on a cat what do you guys think anybody discrete disgrace yeah because those are the you can count them right it's going to be discrete since it is something you count okay okay now um these are about um

6:30 the quantitative variable has a two two breakdown one is the discrete the other one is the continuous okay now the next one is measurement scales okay instead of categorizing the variable into qualitative or you can say quantitative we assign them with the various levels based on their characteristics okay what is those characteristics are these are the four measurements scales are different types of data with each building on one of these blow okay the first one is the nominal nominal is the very variable one that's simply characterized as a name okay categorize or names of the variable like

7:17 a hair color the most general level of the measurement data is just a name or a category right like when we are looking at the data set there is going to be a like a name or category there is no order of any data since there is going to be no numbers you cannot do any arithmetic on the left in this level of the data example for example examples and for this one is going to be gender card name ethnicity and race okay that is one type of a category right there right these are the examples that explains right it's going to be category okay the next one is the ordinal uh or no is the variable categorized but

8:05 has a specific order like a course grade it's a still category but the course grade a b c d they have some kind of meaning okay this is lower little more specific norm than normal variable data that is a normal but you can now put them data in the order since one value is more than the other or more or less than the another value like uh if you talk about the grades right like a grades are a b c d place in a race for second and third size of the drink small medium and large okay those are going to be examples for

8:51 the ordinal and the next one is the interval the interval variable is going to be a quantitative this one is the quantitative that means dealing with the numbers so we have a uh have an order of numbers the difference between these meaning is but um okay the difference between the values have a meaning but a value of zero does not mean the object has a no value the easiest example is going to be the temperature clearly 64 60 degree fahrenheit is more than 30 degree fahrenheit zero degree fahrenheit doesn't mean it's a no temperature it it could be colder right that means in this case

9:40 it's not like a zero does not have a meaning the zero has a meaning here does that make sense right because the zero means it's going to be colder temperature right but as compared to if there is a no meeting okay now that's the the example in this case they are giving us the data is the ordinal that you can now subtract one value from the another that makes sense you can do the automatic one only the addition the examples are the temperature and time on the clock right the time on the clock are going to be the examples the next one is the ratio that is the final and more persistent level of

10:26 measurement is the ratio ratio is the variable has all the properties of interval variable but the ratio of the two values have a meaning a value of a zero means the absence of that quality right that means in this case zero has a no meaning okay a simple example might be point earned on an exam a score 80 is the twice the value of the score of 40. score 0 is clearly mean the student earn no points on the exam right this means the zero has a meaning here right um note the difference uh that grade a versus f where you could just be or no

11:12 data is the interval but now divide one value by another ratio makes sense now you can allow all the informatics on this data example can be in this case height weight distance and a time okay okay now here on the next page there are an examples okay before we go there normal and ordinals are for qualitative variable interval and ratio are for quantitative variables okay here are some examples of these examples the answers are already written for that gender it's going to be normal because it's just that category

11:58 iq it's the larger the iq value employs the uh has a higher iq right zero iq doesn't mean anything that means the individual is dead okay an iq score of 120 isn't means the twice as smarter as the 60 right that means that's why this is the interval okay but to be in this case to be a ratio that means that it like a 120 is going to be is a double of 60 right but in this case it's not okay distance um commuting that's going to be ratio unlike iq it is possible to have the distance commuting uh of a zero okay if you are working

12:45 from home right that means you are staying home just like right now um ratio of two distances have a meaning right that is going to be the ratio pain uh rating that is going to be ordinal because the that means something when it's going to be or no y it seems to be the interval or even ratio variable but that's not the case clearly 7 is more than 5 5 is more than 3 that doesn't means the difference is two for both same cases that means the what is the difference between ten and eight okay because those differences are not consistent right

13:31 that's why this is not an interval variable right that's why you guys have to be careful this is going to be the ordinal okay not the interval okay now we have uh some examples which one we are going to be doing ourselves right here in this case we are going to be talking about um first one is a time of a first class okay in this case what do you guys think is this is going to be the first time interval it is going to be interval

14:26 since it is a number okay now somebody says what uh zero clock means midnight not the absolute like if somebody says zero clock right what is that means that means it's a zero clock means it can be midnight that doesn't mean the absence of a time right okay okay here color variable they all are variable like which one it's a phenomenon it's a nominal that's right yes

15:13 nominal okay since it's not a number and there is no specific order for hair color right that means it's going to be nominal okay length of the time to take a rest couldn't this be the ratio or the interval again this one is a ratio because in this case if you have a like a zero min uh you take a zero minutes to take a rest what does that means you didn't uh take any time to rest right that means it's a ratio okay since it's a number and zero um zero means nothing there right

16:01 the other one has a zero mean something right this one has a zero means nothing right it doesn't take any any time to complete it right like if somebody says i take zero minutes to take a rap test right that means what does that mean they took nothing right that means zero means nothing there okay um next one is the age grouping baby toddler nominal right now is the order yeah because that one has you can write in the order right that order has a meaning like a b c d grade first third third right um that is

16:48 ordinal okay okay now this this is all for our chat section 1.1 is anybody still copying or you guys are good with this okay um now we are going to be set going to section 1.2 okay section 1.2 here we are going to be talking about like to collect the data how we are going to be sampling and what is those sampling means right um like what kind of samplings they are and first one is like uh and like what

17:36 is the samplings are like how those samplings works how we collect the data okay the first one the sampling methods the first sampling method we have is simple random sample okay what is the simple random sample is going to be a sample that is selected from the population in the way to ensure that every one different possible sample size and has a same chance to be selected and also every individual associated with the population has same chance to being selected okay that's like everybody have a chance to be getting selected right like okay if i give you an example

18:24 here okay like because population is bigger group sample is a smaller group okay let's assume population in for this example is of our whole class all 40 students okay i want to pick the sample of five students okay i want to pick a sample of five students if i said i'm going to pick only female students is i'm giving everybody a chance to get picked no no that means then i'm going to be choosing only the female students right if i say i'm going to choose only the male students i'm not giving the chance to everybody right like when

19:09 everybody have equal chance to get picked right nobody knows who's going to get paid okay if i say oh i'm going to look at on my screen i'm going to pick third person no then everybody don't have a chance to get picked only the people that's going to be the third person will get picked right the simple random sample means anybody have a chance to be getting picked in that sample does that make sense guys yes okay now if you look at it this is the picture um like showing uh you guys have on yours i don't know i don't think so they are printing in color in yours this whole box rectangular box

19:54 is a population write these smaller dots if you look at it there are like a dots that's the population what what is the red ones are that means you are randomly picking those people for your sample okay um now what is the what what is how we can do that let's suppose they select the four people from a group of 12 here are some common way to select uh select simple random sample these are like four ways uh they are giving us here write everybody's name on a slip of a paper and draw two from a hat okay right possible sample size two on a slip of a paper draw one from

20:42 the hat number each individual use the technology randomly selected one through 30 right like that means you can do that right that make sure everybody have a chance to get picked instead of uh only a few people okay you guys can read the more like this one right here this is another visualization this is like these are 12 12 people they randomly select people to make a group of four okay they randomly select the group of four okay now uh what do you guys think uh how we will write this one choose a simple random sample describe how you can take a simple

21:27 random sample from your class what do you guys think choose every other person maybe if you choose every other person is everybody have a choice to get picked no yeah no then it's not simple random sample um give everyone a number and then just randomly select two numbers you can sit you're kind of right actually you can give everybody a number and then put all uh write a number on a slip put all those numbers in a basket and like shuffle those numbers and pick randomly like four does that make sense for investment

22:15 and climate change that's kind of what i'm saying for a vote which is likely to be scheduled later this okay uh now uh we can write that down right like what we just talked about [Music] give each student in the class a number okay using a random

23:01 like we can use like technology right random number generator you could then pick the number of students you want to pay okay now

23:47 how not to choose a simple random sample okay you want to choose five students out of 20 give some example that is not random sample i think what deidra said what you said at that time like picking uh oh i said every other student every other student that's not going to be a random sample right yes does that make sense or you can say oh i will pick uh all people in the front row right like on my screen all the people on the top row right uh then they are not then they are not random because then everybody don't have a choice okay you guys can write either example

24:34 i will say choose um choose five students from the [Music] probably we will think about we are sitting in the classroom front row right then um then that's not a simple random sample okay okay that's that next one is sterified okay sterified sampling what is terrify sampling is different with this technique we separated the population using some characteristics then take the proportion random sample

25:22 from each okay okay now like this one was a whole population right here okay what we did we break down the population into smaller groups okay then what we are doing from those smaller groups we are picking randomly few people okay that is called the sterify sampling okay if you want to visualize like um what we like think about like i'm going to give you an example for like our like if you like cuba city they want to survey some people what they do they divide

26:08 the yuba city into four zones right what is those four zones are going to be those groups okay now as you can see they're divided into north south east and west jones right there's four zones they're divided into that means those zones are going to be these groups like that and then they are picking randomly picking people to survey from east east john does that make sense right um and that's going to be like if you look at the visualization i have a two of these one is like this you have a groups like this and then you pick randomly those red dot means those are the ones we are picking

26:54 another visualization is like okay these are different colors right this is one group this is one they can be one smaller or other one bigger like this one has a three this one has a six this one has a three right because these are with the colors and then they will pick randomly from each of these to make a sample okay to make a sample okay the next one um okay here's an examples we can look it up we can highlight this i'm going to use this as a second example or maybe first one too let's read it if you look at the musical preferences you

27:40 can find dividing the groups into age groups you can see that they're dividing into age groups those are going to be the groups conduct simple random sample inside each groups you can calculate the average prices of the textbooks you can divide the individual groups by the majors right they're talking about price on the price i think maybe i have to read individually before okay another example can be using sterify techniques you can be sampling the people at yugo college you make the surface and number of students faculty staff are selected okay that means they divided into groups as a student faculty and staff and then we are randomly picking the people from each group

28:27 okay are we clarifying that these sterify are we everybody's good with starified we divide the population into groups and then we will pick random people from each group when i'm saying random that means every person from that group has a chance to be getting picked okay the next one is the systematic sampling where we have a randomly choose uh starting place select every kid individual okay just like what's deidre gave the example like every other person to pick that's going to be systematic sampling okay you select fifth item

29:13 on the assembly line right every tenth name on the list right every third customer coming to the store that means you have a system right every time you're going to be using that are we good with this systematic that's the easiest one yes okay now like this one is another visualization like they have these 12 people and they are picking every um third person then firstly they pick the second one after that they start picking every third person okay the next one is the cluster sampling okay the cluster sampling is kind of sterified

29:59 uh what we learned but if there is a big difference okay you need to be careful is we break down our population into groups just like what we did in the sterified sampling but what we what is the difference is in this case you break it down into groups but when you're taking the sample you're taking each every person from that group right in this case we break down population into uh called clusters randomly pick some clusters then pull all the individual from the clusters but in this case what we are picking we are picking randomly those groups and then we will survey or what we want

30:46 all the people look at this does they have a few clusters right here and then what they did they picked this cluster they picked this cluster and this cluster as our sample okay now do you guys see a lot of people get confused here okay stir-fry sampling we break it down into groups then we are picking people from some people from each group okay here in the cluster sampling we are breaking them into groups then we will choose whole group whole group to sample okay just like there is another cluster like this they break it down into one and two

31:32 three and four five and six seven and eight nine and ten eleven and twelve and then they pick whole this cluster right here five and six and they pick eleven and twelve okay okay this is important right here many people get confused with the stir-fry sampling and cluster sampling in sterify sampling we use all the groups but from the all the groups we will pick some members in the each group okay from the all the groups we are picking random people right some people in the cluster sampling the way the other way around we pick some groups and all the members from the each group

32:18 okay okay the next one is a convenience sampling which one is one is the researcher pick the individual that is easy to pick for the research to collect like for example if somebody wants to research something about drinking and driving and then they will stand outside the bar and ask people the question about the effects of drinking and driving that means they won't just want to pick like an easy way to collect the data right but that data is not going to give that person the what is the write it okay now here one more definition it's a consensus is what is it is

33:06 when every individual of interest is measure that's called the census okay now what we are going to do we are going to do some examples where what we learn we are going to be using those and see if we understand what kind of sampling is that okay okay now what i'm going to ask i'm going to ask few of you guys to read the problems for me okay and then we will talk about and i'm going to pause the recording and then i will um when i will say the answer and i will um

33:53 unpause it okay now in the a part what they are saying obtain the list of patients who had a surgery and all banner health facilities divide those patients according to the type of surgeries draw the simple random sample from each group what do you guys think stratified it's a sterify because what you're doing whole population you're breaking them into groups what is the groups are what type of surgery they get right then from each surgery you are picking them randomly that's going to be starify sampling okay let's write down that

34:44 okay now this is a starify sampling now here in this case what we can say because they are dividing them into a type of surgery right like because we need to look at what is the what is the issues can be like we got the it's a starified sampling um now because what what they're doing they're trying to figure out what is the co reason was for them to do that to figure out the manager will want to access the incident of complications after the surgeries okay now there are going to be some

35:31 types of surgeries they are going to be more complicated as compared to the other side of surgeries right of course terrified sample would show that on this right but this will show you because you're dividing them into each groups okay the next one who's going to read the next one for me i will pause hey okay what do you guys think in this case because they are saying um they are picking all the patients and labeling them and using randomly those patients what do you guys think what is this is going to be random sample random sample perfect

36:19 okay random sample okay okay i will read the c randomly selected some banner health facilities from each of these seven states then include all the patients on the surgery list of the states cluster that is going to be cluster right because what they are doing randomly selecting the facilities from seven states right um then they're dividing them into seven states and including all the patients from there right that is going to be cluster okay and the next one it is at the beginning of the year instruct

37:04 instruct each banner health facility record complications from every hundredth surgery what do what do you guys think systematic systematic and then instruct each banner facility record any complication from 20 surgeries this week and send the results convenient that is for convenience is it convenience because it says any from record any complications yeah

37:51 no but because they're choosing like 20 20 surgeries from this week that's it instead of looking at like overall data right because there is a possibility there is can be no complications from this week's surgeries and they can be minor surgeries and making more complicated surgeries does that make sense just like uh when i gave you an example like standing outside the bar asking drinking and driving that's the same thing like okay you're instead you don't want to do anything you just pick up this week's 20 surgeries and look at the data okay okay

38:36 and these are like a summary of what we learned like this is like summary is showing the pictures again for you guys to more visualization okay and what we are going to do now before we start section 1.3 we are going to take a break okay let me pause okay now our next section is 1.3 which one is experimented design and in this section we are going to be learning about the how when we talk about the experiments okay before we do that and we are going to be learning about explanatory and response variables okay when we ask the question about

39:22 relationship between the two variables we want to know we sometimes also wants to know determine if change in one variable change the other one consider the following uh rephrasing of the earlier questions from the data if there is increase in the medium household income in the country does that drives the increase in the population okay in this case we are asking whether one variable affect the another if there is underlying belief the median household income is an explanatory variable that will affect the population change

40:07 which one is the response variable okay that means that's this is here the explanatory variable that might affect something and because of that there is going to be another change that is going to be the response variable when we suspect one variable might causes effect on another we if we le first variable is explanatory variable and the second variable is the response variable okay [Music] another way to define that's the definition down here is also explanatory variable is also known as independent or predicted

40:53 variable it explains the variation in the response variable in the experiment study it manipulated by the researcher response variable because it's a depends it depends uh it is a dependent or it's an outcome it is the value is the predicted or the variation is explained by the explanatory variable in the experiment study this is going to be the outcome the measure following the manipulation of the expel explanatory variable okay you guys will see the examples um like uh if we are giving a plant food um because of the plant food

41:39 uh the the height of the plant increases right that means in this case the plant for giving the plant food is going to be the explanatory because of that the response variable is going to be the height of the tree okay okay now um now we are going to be talking about now our main objective in this cash section is going to be the two other common data observational study or design experiment like we will figure out if this is the observational study or this is going to be an experiment um firstly we need to figure out which one is going to be the experiment which

42:24 one is going to be the observational study okay now i'm going to go to the straight to the definitions okay the observational study that measured the correct distance of a population by studying the individual in a sample but does not attempt to manipulate or influence the variable of interest okay that means you're not making any changes you just collect the informations the way it is right for example a good example is going to be pure research center just click on the article um this is the article observation study this one is a sample group and ask them a question like you just

43:09 ask them a question you're not making any changes design experiment which one is an experiment is the explicit to do and close the results then try to attend determine the effect of a particular treatment that has an ah an outcome okay here we're trying to change something okay now here the design experiment is the treatment of the individual refer as experimental units or subject attempt to isolate the effects of the treatment of the response variable okay and a nice example is going to be check on this article for national public radio effect of the exercises on

43:55 the fitness okay now here it's just like same definition it's written again it's when the investigator collected data watching and asking the question he does not change anything this is then that's going to be the observational study okay like for example when you're going to the doctor's office you look at the medical records you're not going to be making any changes you collect the information from the medical records which one is already there it does not make any changes to that okay and an experiment is when the investigator changes the variable are imposed in treatment to determine the effect like for example um like if somebody's

44:41 on diet they give us a special diet then they change their diet and see what is the effects or not on that because of that diet what is the effect on is it's depending on what kind of studies they're doing if there are studies doing on the diabetes right they give them a special diet and then they check what is the effect after that special diet is or they want to lose the weight they are providing a special diet right but they are making imposing something they are giving a treatment that's an experiment okay now here are some examples then we will look at it like if if that's going to be an experiment or observational study okay the first example is the pull

45:27 students to see if they favor increase tuition okay what are you going to be doing uh you're going to be pulling them and asking them a question do you favor increased intuition or not observational that's going to be observational that's going to be an observational study okay and then the next one is give some students a tutor and see if their grade improves desired experiment that's going to be an experiment or you can say design

46:14 experiment okay what we are doing because in this case what is the treatment is going to be the tutors okay um probably is the biggest difference between the observation study and design experiment is the association versus quotation right the observational study does not control any variable the result is the association because the variable controlled by the design experiment can be conclusion the condition all right okay the next one is called confounding variable or you can say it can be lurking variables that's right here

47:09 okay uh sometimes we call the compounding variable also that's another name it's a variable that affect both variable of interest but neither of them known and and is not in claude acknowledged okay for example i'm going to give you an example you guys can read this example what is given here but i'm going to give you an example you guys remember i gave you an example about there was a tree you put the tree uh tree um you give them a fertilizer and then you check um like or some kind of new food to a tree or plant and see

47:56 the what is the height increase okay but we didn't talk about where that plant is if you leave is that plant needs sunlight or like a tree if that tree is in a sunlight versus in the shadow right if the tree is already in the sunlight that can have an effect versus if the tree is in the shade right and that can be a confounding variable or you can say lurking variable the lurking various effect on both variables of interest but either of them are not known or not acknowledged right like that like if the tree is on the shade we didn't acknowledge that that can be a lurking or compounding variable and then you guys can read this one this

48:42 one is there talking about coffee uh health information where you guys can read that information here okay now um we are going to be talking about types of observational studies right i'm just going to give you like definitions very quick right and because most of these the solutions are given and then i will go with the experiments okay um the types of studies first one is going to be the cross-sectional study okay what is the cross-sectional study is going to be the data is observed measured collected at one

49:28 point in a time okay one point at a time and the next one is going to be retro receptive or case control study okay that one is the data is collected from the past using the records interviews and other similar artifacts okay but here we are collecting the data from one at one point in the time here there this one is the from the past okay um here are some examples you guys can read and then we have a prospective uh or cohort study that data is going to be collected in the future from the

50:14 group for sharing common factors okay now here is an example i'm going to be talking about okay one edit one at time the other one was from the past resto or case control was from the past a prospective or cohort study is from the future right okay now let's look at this one here they're talking about this article on bbc and this section described the study concerning the damn nation or midlife ills according to the article research followed the more than 11 000 people over the period of 12 to 14 years they found that they were that the smoking diabetes

51:03 high cholesterol pressure blood pressure were all the factors in there what what type of observational study was this cross sectional case study or cohort okay because study a research track 11 000 patients this one is a cohort study okay because they're going to be cohort study they're going to be collecting the data right for 12 to 14 years right that means they are collected the data for over time right not from the past okay in this one the national institute of environment health funded and study in iowa regarding the possibility relationship between the ran

51:48 um levels of incident of cancer the study gathered the information of 413 patients who had who had 12p cancer and compared those results with 614 a patient who did not have have a cancer what type of study is this one okay this study is going to be retrospective gathering the information about the group of interests those with the cancer comparing with them that control group without cancer this one is going to be case control study okay which one was this one right here okay um um i'm still confused on that first one

52:35 that was cohort why wouldn't it be the cross-sectional study okay cross-country study is like when you collect the data at one point not over time so like it would be like one year or just like one day yeah just like that okay okay now you guys can look at the other examples like this one is the example for um the different examples that all the solutions are written now i'm going to talk about the experiments okay experiment we already talked about right the experiment is going to be

53:20 the definition like you have a going to have a control study which morse treatment are applied like first one is a design experiment okay experiment that you observe and affect the varies of the treatment that on a response right that's going to be the response variable that means you're looking at how that is affecting like as in and give you an examples for the like uh some kind of special diet is given okay these are like okay experimental units the person on object which one at treatment is applied treatment it's a condition applied on the experiment like it can be a diet right response variable is the variable

54:06 of interest like if when i gave you an example about the diet like it's depending on if some i said it can be the uh effects of a special diet on somebody's uh diabetes right then our response variable is going to be how the diabetes change right it can be uh how the special diet how somebody's lose the weight that means in that case the response variable is going to be losing weight right it's depending on what is uh that is the factors the variable which affect the response variable right which one is a we call the explanatory variable okay okay now here this one they're giving

54:51 you an example they're trying to figure out like what is the response variable and treatment is you guys can read that um i'm just going to go a few examples like all this is like for you guys to read i'm just making sure i'm covering what i need to okay i'm going to go over a few things okay first one um like these are like a different types of experiment one is the complete random design um and this one is a completely random design is when each experiment

55:37 unit is assigned to a treatment completely randomly right this is a similar to a simple random sampling right um this one they have like a group of students they make both groups into 30 students but they're randomly picked one group they're giving a treatment a no practice exam the other one is a practice exam and then they're comparing their result okay but how they're doing completely randomly right they're applying the experiment on that right okay the other one is match peer design that's the second one there is a lot more information written you guys can read that okay

56:23 match pair design is the a match pair design is when experiment units are paired up and each period is assigned to a different treatment okay now here in this case all students take a pre-test okay now this is an example and then they will divide into randomly two groups and then they're giving at different treatments okay here what is the difference between this they're these students were divided into two groups and they're giving the treatment one was practice test the other one not and then they compared the results but in this one they give them all these students pre-test and then they randomly assign the two into two

57:09 groups and they one get practice the other one don't and then they compare the results but then they have already the results from before the last one they don't have the results from before then in this case they can check is this woe a fact because there was a already pre-test done okay okay the next one is randomly block design okay and this one a random block design is used when experiment units are divided into a groups called blocks with each block the experiment units are randomly assigned the treatment this is a similarly like sterified

57:54 sampling okay this is a similarly like standardified sampling okay here they're they're saying student merity is a huge factor in college success maturity another factor split the sample by academic year the first year versus the second year right okay they split these students by the academic years then they're giving them group of 15s and then they're giving them those treatments and then they're comparing the results okay the next one we will talk about there is a no matter which experiments you conduct but these are going to be there are going to be you should consider the falling the main ones we

58:40 are going to be there is a application there is a blind study there is a double blind the you guys can read the repetition and experiment uh on one or more subject you can make sure the sampling of is a large enough to distinguish the true effects from the random effects right they just want to check like is that's the results are getting after the repetitions the blind study what is the blind study is individual does not know which treatment they are getting or or they are getting a placebo okay for example right now with the kobe what they were doing when they were testing the vaccines when they were testing the vaccines they

59:27 asked for the people to volunteer to test the vaccine what they did they did like in the first study they would did about 30 000 people but those 30 000 some people they got placebo the other people got the actual vaccine but those people don't know they got an actual vaccine or they got a placebo that means that's the going to be the blind study okay now somebody says what is the difference between double-blind study double-blind study means the people who are administrating these shots they don't even know if they are giving them actual vaccine or they are giving them uh placebo the somebody in the background because that's the real research is right there's somebody nobody tracking like there there's a

1:00:14 numbers on there somebody knows which one is the actual vaccine which one is not yes madison i do have to go okay okay whether the individual or not researcher knows whether they're getting which treatment or which treatment is getting the placebo right this is important to know there is no bias in this individual or the researcher right um like most the time they use this one double blind right to figure out because then there is no bias okay okay this is about section 1.3

1:01:02 these are like more the definitions right like when we are going to be using the definitions and stuff let me pause the recording for a minute okay now in this section section 1.4 again it's more definitions guys you guys can read it more uh carefully i'm just going to go over like few main things and i will talk about the examples okay again you guys have all these definitions written details and i gave the examples a lot like examples are written in detail those examples we need to do i will do it in the class okay first one we talked about lurking and confounding variable in the last section too

1:01:49 which one is you cannot uh rule out the possibility that observed is affect some and other rather than the factor being studied okay an example i i think i gave you this fertilizer example uh when it's given some plants and no fertilizer to other plants but then no plants uh fertilizer plants are placed in the location that does not receive the direct sunlight you won't know the plants that received the fertilizer grew taller because the fertilizer or it's a sunlight make sure you design the experiment to eliminate the effects of compounding variables like controlling all the factors that you can that means you need to make

1:02:34 sure all the other factors are same right like in this case the compounding variables are going to be putting those plants in the sunlight too as compared to the if they are not in the sunlight um the next one it's over generalization where you do study one group and try to say it will be happen on all group and an example doing the cancer treatment on rats just because the treatment works on the rat doesn't means it works on hormones right that means you need to be careful with the overgeneralizations right okay and then cause and effect is where people decide one variable cause the other one just because one

1:03:21 variable is related or correlated unless the study was done an experiment where the variable was controlled and you cannot say one variable cause the another most likely there is going to be another variable that can cause both an example can be relationship between number of drownings at a beach and ice cream sales there is doesn't mean ice cream sale increase the cause of people drowning most cause of uh for both increases can be heat you guys see that is cause and effect okay sampling error the next one is the sampling error which one is the sampling error is the

1:04:07 result from the using the sample to estimate the information regarding the population there is a difference between the sample result and the true population results which one we are going to be all learning okay this is an unavoidable and results in the fact samples uh differ from one from another an example you take a same sample of five people's height in your class you will get a five five numbers if you take another five people as a height in your class you will likely to get five different numbers right there is going to be really nothing we can do about this type of er unless we

1:04:52 sample every single individual in the sample there is will be some error in our results okay that means you need to be careful with that okay um like in this case like we cannot do anything right about some stuff non-sampling errors are the results from the survey process um there is a sample collected poorly either through the bias sampling or the through the error measurements care should be taken to avoid these kind of errors i still don't understand the the sampling error like how does that how is that an error for the height okay now if if i say oh i want to use

1:05:38 five people from your class to figure out the average height of whole class right i want to predict but if i pick five people right and their five peoples their height is going to be different and i pick another five people their height is going to be different their average is going to be different you see that okay yeah yeah that means their average like first five people's average and the second five people's average of height is not going to be same so wouldn't it wouldn't be accurate for the entire class no no what we are saying like when you go from one sample to another sample if you pick five people or 100 people the values are going to change your values are not going to be same

1:06:25 if i say i want to pick 100 people and from the yuba college 100 students and want to find the average age if i go there pick 100 people my average age is going to be different actually you go and pick 100 student a different 100 student and calculate their average that's going to be different because we are not sampling same students does that make sense yes okay okay okay the next one is survey have their own bias can be occurs like few issues can be with the surveys wording the questions

1:07:11 ordering of the questions the manner of the survey conducted and the response of the rate of the surveys okay the the the bigger one is the way the questions are worded okay a lot of time people write the questions they want like predict the the way what they want the answer like particular way you have no choice to give one answers versus the other answers okay okay now look at the wording questions can cause the hidden wires in which the questions are asked the way the person responds a certain way okay and an example can be that a pull was done where the people were asked if they believe

1:07:57 that there should be an amendment to the constitution protecting women's right to choose about 60 percent all people questions said yes another poll done where the people are asked if they believe there should be amendment to the constitution to protect the life of unborn child about 60 percent all the people questioned said yes these two questions deal with the same issue though giving the opposite results but how the questions were worded affect the outcome right because they they were asking yes or no questions right and then you you just people just choose one right that means they are not giving you one that means there was a hidden bias

1:08:42 there same thing the ordering of the questions an example like which order the questions are given an example and this one can be there should be a fine for texting while driving but perceive that question is question asked if you text while drive yeah people are going to be biased here how they respond to that question okay non-response surveys like if we do the surveys and where the service but not everybody return the service you can calculate the response by dividing the number of returns by the number of surveys sent most response rates are

1:09:27 around 30 to 50 percent if the rate response rate is less than 30 percent then it's a poor survey a poor very poor if the results survey results are not valid right that means you cannot be valid if they are less than to reduce the non-response it is better to conduct the surveys in person though very expensive phone service right can be ineffective physical mailing is leased to desirable surveys right but they don't like mostly people don't respond to those the next one is the volunteer response where the people are asked the question by phone or email or online the problem in this one is only people who are really care about

1:10:12 the issue are likely to call or email okay you need to be careful about that okay now here you guys can read a little like there's a lot of details are written here i'm just reading a few things okay the bias in study okay now here what is this one i suppose mathematical department at a community college would like to assess whether the computer waste homework improve students test score they use computer-based homework in a one classroom with one teacher and use the traditional pencil and pen homework in a different classroom with that different teacher the students using the computer based

1:10:59 homework had a higher test score what is wrong with this experiment what do you guys think what is wrong with this experiment is it over generalization yes little bit more thinking i can't to me

1:11:53 okay now to me uh the issue here is um that there can be two different teachers there are two different teachers you don't know if the better test scores are because of the teacher or with the computer system okay to better design that means it need to be same teacher for both classes then you can look at it because that same teacher is teaching same class right with the two different one to the online homework the other one is paper and pencil you guys see what i'm trying to say here yes the scores couldn't they may not

1:12:38 have been related to the materials that may have been related to the teacher yeah that means since here were different teachers um you don't know if the batter has score far because because some somebody's teaching style can be different right because of the teacher

1:13:28 or the computer based homework okay a better design is going to be in this case to have the same teacher for both classes okay a better design would be have same teacher teach both classes okay are we good

1:14:19 okay that means what we want only the difference we want in those two groups to be one getting homework online one is getting homework on a pencil and a pen right sorry on paper that everything else stays the same then it's not going to be the biased study okay the next one in this case is they're saying cinnamon was given to a group of people who had a diabetes then the blood pressure glucose level was measured at a time period all the factors for the pure each person's were kept can same their goku's levels were went down did some cinnamon cause the reduction

1:15:15 what do you guys think the cinnamon is the cause and reduction in the glucose is the effect but like in in this case the question is did the cinnamon cause the reduction what do you guys think about that yes yes because in this case because they are saying they're keeping all the factors are keeping same right um because all the other factors are same right since this was a study where the use of cinnamon

1:16:03 was controlled because they kept everything same right and all other factors were kept constant from person to person than any changes in the glucose

1:16:48 level and the um okay okay there is need to be one more there is a [Music] okay um is the link between spray on tanning product and a lung cancer does

1:17:33 spraying on any products cause the lung cancer what do you guys think on this one this one is not control because in the other one they're saying all the other factors we're saying they're just over journalizing here right they're saying since the only link is not this study is not a control study you cannot say increase the uh tanning is spray is causing the cancer right there you can say there is a link but you cannot say it could cause it right because you cannot sure spray causes the cancer right because it's not all the it's not a control study okay

1:18:19 since there is only a link and not a study controlling the use of tanning spray they cannot and then you cannot

1:19:05 say that increased use causes lung cancer okay one last one this one is they're saying a researcher conduct the study on our use of ibuprofen on humans and find that it's safe does that means it's a safe for all species no that is the over generalization right

1:19:53 you cannot say if one drug is safe on one species that doesn't mean it's safe for all these species right you will say no and you can write that part are you guys good can i go to the next page okay and then the next one is an aspirin has been used for years to bring down the fever in human originally it was tested on white males between age 25 and 40 to be found to be safe is that safe or given to everyone what do you guys think

1:20:40 nothing because it was given only to white males and we cannot say it's not like tested on the children's right we cannot give it to anybody like you cannot say it it's safe for everybody is that also over generalization oh it says it right there okay um this is all about section 1.4 and we are done with the chapter one okay and let me stop