hello everybody now um this video is going to be me going through sampling methods um not experimental design which is what a lot of people actually for some reason get confused with but what we're looking at today is sampling methods and what we mean by sampling methods is how participants are actually recruited for research um it obviously goes without saying that studies have participants and up until this point I'm willing to bet that you haven't really given much thought about how those participants actually came to be in this study which is what this piece of videoness is about so um there are five ways that you need to know about how researchers recruit their participants and we're going to be looking at the different ways of how they do that now before we do that you kind of need to know a little bit about syllables and populations and you know a little bit about these terms they're going to be thrown around now one way to think about it there are three different populations there is the general population okay this is everybody in the world okay fine but when you do a piece of research you typically uh have a Target population in mind so we don't do a piece of research to try explain behavior for everybody we do a piece of research with a Target population in mind and if this is the general population a Target population is basically a much smaller subset of that general population so for example for example um students about in Federal you have a general population which is everybody in the world but then said to want to do a piece of research and I've got a Target population of the 4 800 students that are at Barn Pedro now here's something really important 4 800 students about I can't test all of those firstly not everybody's going to say yes and everybody has a right to say no to the piece of research secondly 4 800 I don't even have time to count to four thousand eight hundred let alone actually tests every every one of those so with those two things in mind what we aim to do is we aim to take a representative sample from that Target population they need to be representative in size they need to be representative in ethnicities and genders and Sexes and um religious beliefs and so on and so forth so as it kind of shows you just here the sample you pull from that Target population you want to be representative but it's that pulling it's that creation of a sample that you're at that we're actually going to be looking at today how do you recruit and collect that sample no I think I've just basically described everything here Target population one thing I didn't talk about and this is why I have it written down is um a consequence of having a representative sample is I can take results from this little gang of people and I can then generalize them out to this gang of people that's the most important thing if you do not have a representative sample you cannot generalize your results to the Target population and that's pretty key to be able to do that otherwise why would you why by the way why would we do a piece of research that can only be generalized to our sample what help does that what help does that give to anybody so these are the five ways these are the five ways so um two are relatively basic the top two both opportunity and volunteer the latter three random systematic and stratified do tend to be more complex they all use random mechanisms and there's a little bit more work to be put into all of them now all of them actually require you to give informed consent right but how you go about that is very different so let's have a look at the first one opportunity sampling opportunities start playing is by far the easiest uh one to do right but there's loads of issues with it so opportunity sampling is also known as convenient sampling and it's basically you asking the people around them just going up to people and be like hello mate do you want to take apartment do you want to take part my research or um going into Southampton City Center or Fair room or Portsmouth or you know Winchester walking up to people like this chugger is doing a charity mugger um and uh asking people if they would like to take part in research now uh what's an issue with this oh where to begin um firstly it's very labor intensive in the sense of you're going to get a high rejection rate secondly and probably most importantly to be honest it's unbelievably biased you are clearly consciously or unconsciously going to approach people who you feel are going to support your piece of research and therefore you're being biased to who is not it's not random is it it's not truly random and you you do need to Define this by the way what we mean by true Randomness is everybody on that High Street has an equal chance of being picked but if you're going up to people and choosing who you go up to by the way you're going to only go up to the Attractive people which on high streets generally there aren't that many you're only going to go up to the people who look like they're going to support your piece of research you're only going to go up to people who you feel would look as if they would actually say yes to your piece of research so it isn't a case that everybody on that High Street has an equal chance of being picked because the ugly people who look like they're going to turn you down straight away are not going to get picked so it's hugely hugely uh problematic and flawed it does save the researcher a lot of time and effort money uh if you ignore the fact that it does take a long time to actually get the participants and it's quite proactive I guess you could say it's quite proactive you know you're getting out there you're you're trying to recruit a lot of people or in your own time and money but like I say it's incredibly biased um and you're going to be unconsciously approaching people who you feel will actually support your original hypothesis so there are in my opinion more weaknesses to strengths for this one the next one though probably not as bad volunteer sampling now I used volunteer sampling for my dissertation you know it's how I collected my participants I stuck a a poster like this up with little things with my number on that you rip off down the bottom and I basically recruited a lot of people that way now volunteer sampling is also known as self-selection and it's done mainly through advertisements formerly it used to be done in newspaper it's not really done anymore it's mostly done through social media now so getting them to fill out survey monkeys or Google forms for example um now the great thing about um the great thing about volunteer sampling is that it requires people to come to you not you to go out to the people it requires people to come to you so what you can do is you can pin up one of these posters advertise it and just let that do the work let people come to you you've got too much to do crack on my crack on with your to-do list so it does require a person to come forward volunteering importantly important to note volunteering does not necessarily mean consent you still have to tell them what is going on um then they give you informed concern but I guess one thing to keep in mind is the type of person who is going to come forward now let me briefly tell you what happened to me when I was 18. I heard a radio that was a thing that used to play music um that I had a radio advert um looking for males aged 18 to 25 who did not have asthma and were generally very healthy to come forward for an experiment I writing it I really I ranging up and found out it was something called flu camp now flu Camp is where you are injected with the flu on purpose um and you are injected with base you quarantine for two weeks before quarantining was called before covert made it uh cool and you basically then got injected with experimental drugs to see if they could actually fix the flu one question I would want you to consider and I may ask this in class is would you would you put yourself forward for that now most people say no they wouldn't oh by the way you got to pay two grand for it you got paid 2 000 pounds for two weeks that's more than I get paid now by the way um for for roughly what I get paid for four weeks um probably shouldn't told you that um uh would you would you put yourself forward for most people say no a certain type of person says yes and this is what we call the volunteer personality it's someone who is more likely going to put themselves forward for tests now my question to you would be does this person represent everybody Noah only represents a certain type of person therefore if we go back to if we go back to all this the sample we are recruiting does not accurately represent the target population because only certain people are putting themselves forward and thus not representing the whole population so a problem with volunteer sampling is what we'll get to in a second but a strength smoothly done is that it could be arguably less time consuming because all you do is you put up a poster and it comes forward another strength by the way is people tend to be quite um people tend to be quite willing to take part once they have volunteered so they already have the time you know almost everybody that comes to you probably has the time to take part in the study otherwise they wouldn't have volunteered so the dropout rate tend or rather the attrition rate tends to be quite small in volunteer sampling which is fantastic but as I said it does actually attract a certain type of person which does lower the religion of sorry does uh lower the validity of the study as well as the population validity because you're not getting every single not getting every single person now oh and it's incredibly biased by the way but yeah Less on that now the next three this is quite important to write down the next three all use random mechanisms and what we mean by that is everybody in your target population has an equal chance of being picked so the next three random systematic and stratified are not biased but they are not biased in who they are recruiting and everybody has an equal chance of being picked so the best way to go about random ones um the best way to go about random is that actually you you talk about the Hat method you're talking about the Hat method now what you do is you for these three uh for these next three ones really important you you have to have the names of everybody in the Target population you kind of have to have a register of some kind of everybody in the Target population so that could be I don't know people are schizophrenia in Southampton you get that from Southampton city council for example or from hospital Wards or doctor surgeries you kind of look for people with Autism in the F4 for example in the Ferrum area and again you'll go around the schools and you'll get a list of everybody and you'll create this big Target population uh people over 70 who live in Portsmouth who used to work in the construction sector for example right a really Niche type you have to get everybody there then all you do is you have all of their names on separate pieces of paper you put them into a container of some kind you pick them out you approach them to see if they want to take part in your research and if they do boom you've recruited someone that is the Hat method now the reason why this is not biased is because everybody everybody has an equal chance of being picked and therefore you are not biasing or prejudicing the sample so what I would save to do with this by the way is this is a very very very common question it comes up most exams uh about how you how would you do how would you do this it's in my opinion one of the most uh common questions that comes up quite a lot now the good thing about this is that there's no research or bias right the buyer the researcher is not consciously or unconsciously influencing who is going in the sample is totally up to chance and everybody has an equal chance of being picked but it's very time consuming um not only do you have to pick an email that doesn't mean they're in this study by the way that's really really crucial just because you pay their name out doesn't mean they're in their study when you've got their name you then have to go to them and say do you want to take part my research if they say no but right Tim Roth I'm going to go up to that person for example so it is actually quite quite time consuming now systematic is also very very similar it's very similar as before you still need a full register of everyone you want to recruit but this time what you were going to do you have all the names in the Hat for example this time you're going to pick systematically so instead of picking every one hour ago this one this one this one this one you're going to do it every fifth time so you're wrong it's every nth number systematic is every nth number you roll a dice it comes up with a three you then pick every third person and approach those to be honest with you it's not particularly different from random but um again there's no research bias it is incredibly representative but it can take quite a while to do now the thing that really irritates me most about sampling methods is people see every nth number and for some reason they get it confused with stratified sampling systematic is to do something systematically is to do it consistently and equally every nth number stratified sampling is what to do with Stratus which are subgroups something completely different so stratified sampling it's very simple even though students over complicate it stratified sampling is about making your sample right there's a small sample the same proportions [Music] as the target population for example generally speaking in psychology the population is 70 female and 30 male therefore it would not make any sense for me to come up with a sample 50 50. because in that case the males would be overrepresented and the females would be underrepresented so my example in this case has to be proportionately similar to that of the target population which is 70 and 30 percent so say I had uh I don't know um I'm looking to recruit 100 people that would mean that 70 people would have to be females a 30 would have to be or 30 would be male in this case so how I would go about that is I would recruit say seven people and then recruit three people recruit recruit seven females then recruit three males recruit seven females then recruit for three males so essentially what you're doing is you are um you are finding out what your stratas are right that's your subgroups what are the proportions of your target population and that that's that's um that's something a researcher will actually try a certain researcher will try to find out the stratus of a Target population so for example for example of the top of my head we talked about people with schizophrenia living in in in Southampton one thing I would have to find out is how many of those are males and how many of those are females because apparently schizophrenia is roughly two to one in terms of male to female there's a the jury's out on it but if my samples or female that sample does not represent the target population so another example here is uh Barton purple students in March so if I were to recruit but a bevel students in March 16 16 year olds are roughly 35 of the population 17 year olds are roughly 50 whereas 18 year olds are roughly uh 15 so if I'm going to recruit a hundred people from Barton Pepperell in March 35 need to be six of 35 year olds need to be sick 35 percent need to be 16 year olds 50 need to be 17 year olds and 15 need to be 18 year olds for example for my sample to accurately measure you guys I said 150 there so a bit of maths so um these These are that's it pretty much um that's it um these are incredibly common incredibly common questions so I do actually have a exam example here ignore that question on the right because I don't know what I was thinking when I wrote it but I would like you please to read through this spec have a look for what the sampling method is this one's actually quite easy to be honest um because it literally does say it but it's your job to be able to pick it out so part of your job in the exam is for you to actually because then they don't usually tell you this easily um part your job is to ascertain what sampling method is used and that's it we've got a lot of experiment uh research method stuff coming up so uh the next go on to next video please which I imagine will be experimental designs