YT YouTube2Text

YouTube2Text → Transcripts → Non Random Sampling

Non Random Sampling — Transcript

by MCO-3 [RM&SA] · 6,454 words · 1,133 segments · language en · Watch on YouTube

Open in transcriber Summarize with AI Download SRT/VTT

Full transcript

0:00[Music]
0:08[Music]
0:14hello learners
0:15i am dr sabot sharwani working with
0:17indar gandhi national open university
0:19the topic which i am going to talk today
0:20is method of sampling part two
0:22which revolves around non-random
0:24sampling i think in our preceding two
0:26sessions
0:27if you recalculate yourself or you know
0:29visit the
0:30facebook live link you will find out
0:33that we
0:33the first session was you know talking
0:35about the sampling and when we have
0:36talked about the sampling we talk about
0:38how the sampling is
0:39quite important when we talk about the
0:41research when we talk about you know
0:42collection of data or when we talk about
0:44you know interacting with the
0:45respondents because
0:47there are certain mythologies which are
0:48going to play a very important role and
0:50then afterwards we have we have talked
0:52about you know the methods of sampling
0:54and talk about you know the first part
0:56which talks about the random sampling or
0:58rather we can say the probability
0:59sampling
1:00so in this probability sampling we have
1:02we have covered you know the various
1:03aspects
1:04that how the probability sampling
1:05differs from non-probability sampling
1:07or you know the random sampling differs
1:09from the non-random sampling so today's
1:11session you know talks about the
1:12non-random sampling
1:13which is also one of the important
1:15ingredients as far as you know the
1:16collection of data is concerned as far
1:18as the sampling is concerned so
1:20anyway prior to going into the depth of
1:21this topic let's you know quickly
1:23recapitulate ourselves what we have done
1:25in our preceding session
1:27and what exactly you know the points
1:28which had been taken care so that
1:30you know the blocks can be built and
1:32gradually we can
1:33mug up you know the the concept related
1:36to research or the research methodology
1:37or the statistical analysis so
1:40the first the very first lecture was
1:42talking about you know the curtain laser
1:43to research
1:44which usually talks about what exactly
1:46the research is and how research
1:49is going to play a very important role
1:50as far as the academicians are concerned
1:52as far as the academia is concerned as
1:54far as
1:54the corporate world is concerned or you
1:56know the present trading commerce is
1:57concerned
1:58because everywhere research is coming up
2:00in a big way and we have observed that
2:03there was a time when research was
2:04confined to to the university systems or
2:07to the
2:08academic institutions or the research
2:09institution but now it had crosses the
2:11boundaries
2:11and everywhere you find out even the
2:14most omnipresent thing which
2:16which the corporates are doing right now
2:17it's doing the research of the consumers
2:19or knowing about the behavior about the
2:21consumers so
2:23we are not going more into the depth of
2:25it because already we have a very
2:26thought provoking sessions which can uh
2:29provide a link to the
2:31learners to go through into that so then
2:33we have talked about the research
2:34thought
2:34how research thoughts are coming and how
2:37on the basis of you know the research
2:38thought you can build your
2:40future plan of action of research then
2:42there are certain faqs which are going
2:44to play a very important role
2:45like the frequently asked questions at
2:48that when we are going to the research
2:49when we are going to choose the title
2:51or when we are going to formulate the
2:53objectives or what could be the modus
2:54operandi behind that so this whole
2:56research methods talks about that then
2:58we have research plan research problem
2:59because
3:00while starting your research you know
3:01the certain problems comes into picture
3:03and so the planning is going to play a
3:05very important role and if you have
3:06planned the things in a very systematic
3:08manner i think
3:09up to some extent you can resolve your
3:11way of doing the things so
3:12then we have formulation of objectives
3:15how the objectives you know
3:16can be framed then hypothesis was there
3:18because this hypothesis is quite
3:20important because when you're going to
3:21into the depth of the research i think
3:23the the first building block is is
3:25they're used to
3:26is to you know having a null hypothesis
3:28or then you can
3:29work on the basis of these hypotheses
3:31which can extract from the objectives
3:33which we used to do so
3:34research design is there then collection
3:36of data is there which usually talks
3:37about you know sample
3:39uh secondary data and primary data and
3:41then sampling is there so
3:42in our last to last session we have
3:44talked about the sampling
3:46and the the very last session was random
3:49sampling so
3:50probability sampling is there so now
3:53uh the very important thing is that why
3:55we go for uh sampling this we have
3:57already talked about this is a very
3:58important aspect because universal and
4:00population is a very gigantic shape
4:02and it's really cumbersome for the for
4:04the individual for the researchers to
4:06you know
4:06go into the depth of all the population
4:08or the universal and come to the
4:09conclusion so
4:10that is where you know the the concept
4:12of the sampling comes
4:14and the collection of sample comes into
4:15picture so from there you know the
4:17things had need to be come up
4:18and then there are certain methodologies
4:21which are
4:22very important and when you talk about
4:23the random sampling we have already
4:25talked about you know the random
4:26sampling
4:26and we have seen that how the
4:28bifurcation of random sampling is there
4:30because
4:31uh the very important thing is that the
4:34that random sampling talks about you
4:35know the sampling outlier
4:37outlier is is a very important
4:39ingredient because
4:40when we study the statistics when we go
4:43into the depth of measure of central
4:45tendency
4:46there you know we have a very stereotype
4:48methods that is uh
4:50mean median and mode and these mean
4:53median modes
4:54lacks you know when the outlier comes
4:56because when you go for the average
4:58average in general you know are in
5:00certain intervals or you know
5:02link to one another but sometimes what
5:04happens when you take average sometimes
5:06you know the figure comes like 5 10 15
5:08and sometimes comes at 90.
5:09if you take average of all three i think
5:12the averages could be somewhat
5:13differs whereas if you so that that 90
5:16is nothing but
5:17outlier because it is different from
5:19from the other other sample numbers
5:20which you have taken so
5:22then you have then you are going to
5:24apply the measure of central
5:26standard deviation or measure or
5:28dispersion or other way of doing things
5:29so
5:30that is the model which is there where
5:32when you are emphasizing more on the
5:33random sampling i think the random
5:35sampling always emphasize on
5:37sampling outliers which is somewhat
5:39lacking so that is the reason you know
5:40we have to come out from there
5:42and we have to apply this
5:44non-probability sampling so anyway
5:47we are just going to throw a light on
5:49what exactly the non-random sampling is
5:51because
5:52when you talk about the sampling i think
5:54is a study of research involves a large
5:55number of population we have already
5:56talked about
5:57and you see universe is at the at the
6:00top then you have sensor then your
6:01sample population then
6:02sample frame and then from here from
6:05here you are going to
6:07take the elements or collect the data so
6:10these are the points which we have
6:11already covered in our preceding
6:12sessions
6:13and what could be the sample is what are
6:16the essentials of good sample then
6:17random sampling methods
6:18and and today we are going just focus on
6:22non-probability
6:23methods so when you are going to talk
6:25about the
6:26probability methods i think we have
6:28already talked about it's a random
6:29sampling does not involve human judgment
6:32requires sampling frame and the
6:35the you know the methodologies are
6:37systematic stratified cluster
6:39or you know the simple random sampling
6:41whereas when you talk about the non
6:43probability methods it is you know the
6:45it involves human judgment
6:47and it has sampling frame required and
6:49it is you know called the quota sampling
6:51convenience sampling judgment sampling
6:53or you know this snowball
6:56sampling so uh so we are going to
6:59gradually talk about and distinguish how
7:00this
7:01probability sampling refer differs from
7:04non-probability sampling because
7:05when you are going to talk about the
7:06probability sampling i think if you see
7:08this particular image
7:09this this probability sampling is quite
7:12scattered in nature whereas
7:13non-probability sampling is quite
7:14purposive in nature
7:16and and it is in a in a very systematic
7:18manner you are going to take so somewhat
7:20you know the
7:20model is like that whereas when you are
7:22going to talk about the probability
7:23sampling you are not
7:25biased whereas when you are going to
7:26talk about the uh this non-probability
7:28sampling you are very much biased
7:30so this is a thing which is there and
7:33now
7:34we have got a very very important
7:35comparison which can distinguish
7:38that what is the difference between what
7:39is the actual difference between you
7:40know
7:41the probability sampling and
7:42non-probability sampling and that is
7:44quite important because once you are
7:45going to depth of
7:46non probability sampling there must be
7:48certain logic there must be certain
7:50justification
7:51if you are not going to provide the
7:53justification i think some way you are
7:55going to lack because
7:56because what happened when we when we
7:58are in the process of
7:59collection of of data or taking the
8:02sample or extracting the sample
8:03i think this this phenomena need to be
8:06understand in a very
8:08contemporary manner so if you see this
8:11probability sampling is a sampling
8:12technique in which the subjects of the
8:14population get an equal opportunity to
8:16be selected
8:17as a representative sample we have seen
8:18in the cluster sampling we have seen in
8:21the you know
8:21the stratified random sampling that we
8:24make certain strata we make certain
8:26group
8:26and from that group you know some of the
8:28representation need to be done
8:30so there must be a model where equal
8:32opportunities need to be given
8:34and there must be a representative from
8:36e sample so that is the beauty of
8:38probability sampling which we have
8:40studied elaborately in our preceding
8:41session
8:42and if you talk about non-probability
8:44sampling i think non-probability
8:45sampling is a method of sampling wherein
8:47it is known that which individuals from
8:50the population will be selected as
8:51sample because
8:52so that could be the reason you know if
8:54you talk about probability sampling it
8:55could be
8:56not biased whereas non-probability
8:58sampling is quite biased because you
8:59talk about you know the convenient
9:00convenience sampling
9:02or you know judgmental sampling or the
9:03quota sampling
9:05or even the snowball sampling or
9:06purposive sampling so everywhere you
9:08observe that you know you are quite
9:10familiar that you are going to collect
9:12the data from from this manner
9:13and and that from this particular class
9:15increase so you are quite biased in that
9:18and if you talk about the convenience
9:20sampling
9:21you are you know for your convenience
9:23like you know
9:24you are doing your research in in delhi
9:27or in
9:27particular state so you have gone for
9:29the convenience sampling because you are
9:31reciting to that particular state and it
9:32would be easier for you to collect the
9:34data
9:34or collect the sample so that could be
9:36the reason you are emphasizing more on
9:38convenience sampling so and and some
9:41certain times
9:42when we go more into the depth of the
9:43judgmental sampling or the snowball
9:45sampling
9:46which is quite talking about
9:49cross layer model where the reference is
9:52going to play a very important role so
9:54that is the thing which is there so
9:55alternatively known as random sampling
9:57this is known as non random sampling
9:59and basis of selection is randomly that
10:00is very true and
10:02randomly you are going to select the
10:04content or you know the sample from pro
10:07if you're talking about the probability
10:08sampling but if you are talking about
10:10non-random probability sampling or non
10:11random sampling you are arbitrarily
10:13taking the
10:14sample so opportunity of selection is
10:16fixed and known that is very true and
10:18not specified in unknown
10:20so research is conclusive it's very true
10:22and end of the result what you are going
10:23to get
10:24you are going to get a very conclusive
10:25result which is which is quite holistic
10:27in nature whereas
10:28whereas when you are talking about
10:30non-probability sampling it is quite
10:31exploratory
10:32we will see there are certain methods
10:34like snowball
10:35uh nonven sampling methods which
10:39which purely based on the reference like
10:41we will send
10:42the questionnaire to respondents and ask
10:45the respondents to
10:46provide a certain you know the lead and
10:48on the basis of that lead we can again
10:50you know
10:51hit that particular respondents and ask
10:53certain questions from them so this
10:55could be the modus operandi
10:56which talks about that how the
10:58exploratory studies there and these are
10:59quite consuming you know
11:01and it's quite holistic in nature and
11:03and takes into consideration you know
11:05the time factor and other things so
11:07results are unbiased
11:08and results are biased because you are
11:11quite
11:11the example which i have quoted about
11:13you know the convenience sampling that
11:14you are
11:15for your convenience you are going and
11:17and choosing the particular judicials
11:19and you know that if you go for other
11:21jurisdictions maybe some different
11:22results will come but
11:23it's really cumbersome for you to you
11:25know collect the data from different
11:26states so since you have
11:28confined yourself to the particular
11:31judicion i think you are quite biased in
11:33that manner so
11:34this is the model which is there and
11:37methods are objective
11:39that is very true you are quite object
11:41oriented whereas
11:42methods are subject-oriented and you
11:44could be quite subjective
11:46and as far as the inferential inferences
11:49is concerned as far as
11:50you know the thing is concerned this
11:53probability sampling talks about the
11:54statistical way of doing the things and
11:57non-probability sampling talks about the
11:59analytical way of doing that thing so
12:01here you know the hypothesis is need to
12:03be tested when you are going to talk
12:04about the probability sampling
12:05whereas here the hypothesis need to be
12:07generated so sometimes what happen
12:09uh when you are you know uh applying the
12:12non-probability sampling
12:13you get certain hypotheses which need to
12:15be generated or it could not be
12:18it could not be an alternative
12:19hypothesis but it can be a new kind of
12:21hypothesis which can again added in this
12:23list of hypotheses so
12:26the impact of these kind of
12:28non-probability sampling is that
12:30somewhere after certain time what you
12:32observe that that
12:34that your research could be lingered
12:36down some in some of the cases so anyway
12:38we have very very elaborately compared
12:41the term called
12:42this particular thing so non-probability
12:44or non-random sampling is a process of
12:46selecting sampling from a population
12:48without using statistical probability
12:50theory and
12:52if you talk about you know the
12:53non-probability sampling theory each
12:55element of member of the population does
12:56not have an equal chance
12:58of being included in the sample and the
12:59researchers cannot estimate the error
13:01cost by not collecting data
13:03from all elements member of the
13:05population so that is where which is
13:06which is there and if you go more into
13:08the depth of this non-probability
13:10sampling i think
13:10it's quite purposive in nature and that
13:12has to do with sampling
13:14addressing the question for research and
13:16the research sample is qualitative type
13:18and the probability sampling focus on
13:21external value for transferring of
13:22issues
13:23and the very important thing is that the
13:26series and control of sample is
13:27non-probability random sampling
13:28typically
13:29smaller than 30 cases again so in every
13:32junction you will observe that you are
13:33quite biased and
13:34you are quite you know making the things
13:36as per your convenience as per your need
13:38so that that could be the reason you
13:40know the result which used to come
13:43it have a very short-sighted approach
13:45and not
13:46giving the very holistic view or the or
13:48the robust report so
13:49non probability sampling have numerous
13:51sample of size which is determined by
13:53the purpose of particular x
13:55component and and this is you know
13:59using more rigid side estimation
14:00procedures so when we are going to talk
14:02about you know the certain types of
14:04these judgmental convenience purposes or
14:06snowball
14:07or you know the quota sampling we will
14:09see that how this non-random sampling
14:12differs from from the random sampling
14:15and what are the usp behind that so the
14:17methods are non-probability random
14:18sampling use
14:19wide range of sampling techniques so the
14:21reason behind is quite clear that
14:24sometimes what happen we are not in a
14:25position to collect the data
14:27or not you know getting an opportunity
14:30that respondents are
14:31reciprocating to us in that manner so
14:34what exactly we are going to do
14:35we are applying the certain methods
14:38which are quite important so
14:39in general we have four important
14:42ingredients which are
14:43part of non-probability sampling that is
14:45convenient sampling judgmental sampling
14:47quota sampling and snowball sampling so
14:51ah if we if we start with you know the
14:53non-probability sampling
14:55i think that this convenience sampling
14:58is talks about in the members of
14:59populations which are selected that are
15:00easily available for study
15:02and if you talk about the quota sampling
15:03sample generated by dividing the
15:05population to separate groups
15:06and selective members from each group
15:08that is non-representative quota
15:10and like you know we have taken we are
15:12going to take a data off
15:13of the mail which is of 50 years or you
15:16know taking the
15:17data of females of of of between you
15:20know 24 to 30 years so
15:22we have restrict ourselves so that could
15:23be the model by which you know the quota
15:25sample used to work
15:26and when you are going to talk about the
15:27snowball sample this sample in which
15:29members of the sample select further
15:31members for inclusion in the sample so
15:32here what happened we first float the
15:34information
15:35float the questionnaire to respondents
15:38and then ask the respondents to give us
15:39certain
15:40references so this is this works with
15:43the
15:44in a model called multi-level marketing
15:46or network marketing or
15:47or you know the reference uh procedures
15:50where
15:51the if we interact with one person the
15:52other person can give the reference of
15:54three
15:54and this could be the model so while you
15:56know elaborately talking about this
15:57snowball sample we will
15:59very easily view that how the things are
16:01going to be changing self-selecting
16:02purposive sample is there which in which
16:04member of population
16:06select themselves for inclusion in the
16:07sample so anyway i think
16:10this is another difference which talks
16:12about
16:13that how the things are going to be
16:14changed and there are certain advantages
16:16there are certain
16:18cost and degree and there are certain
16:19disadvantages also so if you talk about
16:22convenience is very low cost
16:24extensively used if you talk about
16:26judgment
16:27it's moderate cost average use and if
16:29you talk about quota moderate
16:30cost very extra extensively use and low
16:33cost use in special situation that for
16:35the snowball
16:36model is concerned so we start with the
16:38purpose of sampling
16:39and we will cover you know all the five
16:41heads which comes under the ambit of
16:44methods of sampling as far as the non
16:45random sampling is concerned we see that
16:47this purpose is sampling is quite biased
16:49in nature and
16:51from the very beginning you know you
16:52observe that that deliberately
16:54the person which had been taken
16:56extracted or you know the sample which
16:58need to be taken is quite biased so
17:00this is called purposive sampling we
17:01have we have populations and purposes of
17:03sampling is
17:04that we are biased that we are going to
17:06take only you know the black color
17:08uh person as a sample so we have taken
17:10only that only so this is the model
17:12which is there if you talk about
17:13snowball sampling this is a chain
17:15referral sampling and
17:16if you see this chain reference sampling
17:18this this chain referrals
17:20works in the manner which i am talking
17:22about that the multi-level marketing or
17:23the
17:23network marketing where the chain is
17:25there so while you know
17:27floating the sample to one person now we
17:30have taken a reference of
17:32two person from this person and these
17:34two person had given a difference of
17:36three more so one person had given a
17:37reference of three other had given a
17:39reference of three
17:40now from that three we have got you know
17:42the huge number of samples
17:44so these are considered to be a mouth
17:46marketing model or you know the way
17:48where we have we have seen and how the
17:50things are going to be
17:51moved and snowball sampling or chain
17:54reference sampling is defined as a non
17:55probability sampling techniques
17:57and which have crates that are rare to
17:59find and
18:00the this sample technique in which
18:02existing subjects provide refers to
18:04recruit samples
18:05so and this is sometimes you know quite
18:08important and this sampling method
18:09involves a primary data source
18:11nominating other potential data source
18:13so what exactly we are doing that when
18:15we formulate objectives when we
18:16on the basis objective we frame the
18:18hypothesis and on the basis of
18:20you know this hypothesis we
18:24frame the questionnaires and then these
18:26questionnaires are sent to the
18:27respondents
18:28and once the respondents reciprocates
18:31then what
18:32exactly we are going to do we can
18:35know we can very easily know that that
18:37you know the
18:38reciprocation is very less the
18:40respondents are not reciprocating
18:42the manner they're supposed to do so
18:44what exactly we are
18:46going to do in that case we take a
18:47reference from from these
18:49you know the the respondents that why
18:51not you can give us some more reference
18:53so
18:53and this nominating other potential data
18:55source is there and
18:56and this snowball sampling method is
18:58purely based on reference
19:00that is very true and uh it either it
19:02could be and these referrals are are you
19:04know
19:05uh are coming in due course of you know
19:07the collection of primary data it's not
19:08like that
19:09that you're going to collect the data so
19:11the the respondents the the bona fide
19:13respondents to whom you have floated the
19:15questionnaires
19:16they can give the the the references
19:19or they can give the referrals so these
19:20referrals are are you know
19:22and when the when you are making a
19:24question here you should you know design
19:25a question in such a manner
19:27that in this this particular question
19:28here you are going to ask
19:30you know the referrals so this snowball
19:33sampling is a popular business study
19:34method
19:35and which is which is going on the way
19:37things are going to be done because
19:38the way e-commerce companies are going
19:40to do the research or
19:42you know the the other
19:45way of research is coming out we have
19:48observed that this snowball sampling
19:49method extensively used where population
19:51is unknown and rare and it is tough to
19:53choose subject to assemble them as a
19:54sample for research so
19:56then we have observed that this
19:57convenience sampling is
19:59is is a very is a different method which
20:01talks about collection
20:02collecting data quickly and fewer rules
20:06to follow
20:07and this is again you know inexpensive
20:09methodology
20:10and easy to do research and this if you
20:13talk about the strength and weakness
20:15because when we are going to do the
20:16sampling
20:17we always do the swot analysis that that
20:18is strength weakness opportunity and
20:20threat
20:21and when we check these two parameters i
20:23think strength
20:24this is low cost time and administration
20:26high participation and generalization
20:27because
20:28uh because you know as far as the
20:30convenience sampling is concerned it's
20:31quite convenient you
20:32you can define the jourdains you can
20:34define your population and sometimes you
20:36can define your sample also
20:38so what exactly we um what what exactly
20:40we mean to say
20:41if you talk about the weakness it's
20:43generalizing subjects id population
20:45results depend on the characteristics of
20:47sample so
20:49this convenience sampling is defined as
20:50a method adopted by researchers
20:52where they collect market research data
20:54from a conveniently available pool of
20:56respondents so
20:57you have got a pool of respondents and
20:59you can very easily you know
21:01as per your convenience as per your you
21:02know pace you can collect the data
21:04so it is most common use sampling
21:06techniques as it is incredible prompt
21:08uncomplicated and economical so
21:10researchers use convenience sampling in
21:12situations where additional inputs are
21:13not necessary for the principal research
21:15and the components of the population are
21:17eligible and dependent on the
21:18researchers proximity to get involved in
21:20the sample so we have seen that
21:22that as far as the convenience sampling
21:24is concerned we have seen that how the
21:25researcher had
21:26had taken no purple figures in the
21:29sample so he had just taken you know
21:31only the sample
21:32which is which is away from the purple
21:34colors
21:35so this is a convenient sampling so
21:36again what i mean to say when you are
21:38going to talk about the non-random
21:39sampling it is quite biased in nature
21:41because this is you know somewhat
21:44talking about
21:44the convenience so select any member of
21:46the population work conveniently and
21:48readily available
21:49and then there are certain applications
21:51of convenience sampling so convenience
21:52sampling is applied
21:53by brands and organization to measure
21:55their perception of the image in the
21:57market
21:58that is very true because they can they
22:00can know about you know the brand
22:01loyalty they can know about the usage or
22:03the behavior of the consumers
22:05vis-a-vis to particular product so data
22:07is collected from potential customers to
22:08understand
22:09specific issues so there is always a
22:12chance that randomly selected population
22:14may not accurately represent the
22:15population of interest
22:17thus increasing the chance of bias so so
22:19anyway i think
22:21if you talk about you know the
22:22judgmental sampling which is again you
22:24know the
22:25the fourth ingredient of of this
22:26particular methods of sampling
22:29which we usually call non-random
22:31sampling i think this consumes minimum
22:33time for execution
22:35and directly approachable to respondents
22:37because
22:38and almost real time results are coming
22:40up so
22:41that is the reason you know you observe
22:42that when any e-commerce company sold
22:44the product
22:45they just put the budge on there on
22:47their portal and the
22:49the moment this consumers log in the
22:52particular portal
22:54or do the e-commerce the it asks for you
22:56know what your experience related to the
22:57particular
22:58product which you have ordered you know
23:00or two days back so
23:02in that case what we observe that this
23:03is considered to be a
23:05going to give the real time results and
23:07easy to conduct it quite easy to conduct
23:09because
23:10uh because the moment consumer login i
23:12think the first
23:13you know the budge or the question comes
23:15uh from the e-commerce portal to the
23:17consumers that
23:18that how you are going to how what do
23:19you think about the product and
23:21how we can you know enhance our services
23:24or other models so
23:25this is the thing which is there's
23:26another form of convenience sampling
23:27where participants are handpicked
23:29from the accessible population
23:31researchers select participants that are
23:33representative of the entire population
23:35and very subjective sample method can be
23:37biased so judgmental sampling is a form
23:39of convenience sampling or we can say in
23:40which the population elements are
23:42selected
23:42and based on the judgment of the
23:44researchers so sometimes again you know
23:47this is quite biased and when you are
23:49going to compare with the random
23:50sampling because
23:51the in this case also they as far as the
23:54final
23:54opinion is going to be considered the
23:56researcher role is very important and he
23:58is going to
23:58do the sampling as per the up its
24:01convenience and as per his judgment
24:03so test markets purchase engineers
24:04selected industrial market research
24:06expert witness used in court so these
24:09are considered to be the judgmental
24:10sampling and then you have
24:12got a quota sampling so if you talk
24:13about you know the quota sampling
24:15this it had also certain strength and
24:18weakness
24:18and its low cost time and administration
24:22is involved high participation in
24:24generalization that is very true
24:25and more representative of individual
24:27population the example which i have
24:28quoted about
24:29you know choosing you know the teenagers
24:32who are who are you know between in the
24:34bracket of
24:3515 to 18 years or if you talk about
24:38weakness generalizing subjects id
24:40population results depend on
24:41characteristic
24:42sample more time consuming so one one
24:45type of this is
24:46another type of non-probability sampling
24:47and must begin with the matrix of target
24:49population characteristic that is
24:51percentage males percentage females uh
24:54ses etc
24:55the set quota percentage of each
24:57character is fulfilled in the
24:58sample so there are certain advantages
25:01again because quota sampling means to
25:02take a very tailored sample
25:04and easy to administer fast to create
25:07and complete
25:08inexpensive and takes into account
25:10population proportions if desired
25:12and can be used if probability sampling
25:14techniques are not possible
25:16so what we observe that sometimes what
25:18happens there are certain probability
25:19sampling
25:20which which is not you know applicable
25:22so in that case you know the quota
25:23sampling is going to be done
25:25so this quota sampling is
25:28taken massively by by the by the
25:30governments and selection is not random
25:32selection
25:33bias poses a problem for example you
25:34might avoid choosing people
25:36who live farther away or people in rough
25:37neighborhoods this may take the results
25:39unrepresentative of the
25:41population so this is a this is a model
25:43of quota sampling where you have seen
25:44that
25:45from this particular group from this
25:46particular uh thing
25:48you know the the male population had
25:51been
25:52taken and who are in the bracket of 50
25:55years so
25:56anyway i think i have talked a lot about
25:59this random sampling i have talked a lot
26:02about
26:03this non-random sampling and this and
26:06i have very you know very ah
26:09meticulously
26:11explained in the very beginning that
26:12sampling is a technique you know
26:14and it's not like that you can you can
26:16just you know uh
26:18consider the population or you know
26:20earmark the population or universal then
26:23just extract the data from that
26:24particular population there must be
26:26certain modus operandi
26:27there must be certain procedures there
26:29must be certain methods
26:30and when you are going to talk about the
26:32methods i think you have to
26:34first see what exactly the sampling is
26:36and how you're going to collect the
26:37sample and what could be the modus
26:39operandi you are going to do
26:40and what are the ways you are going to
26:41apply like you know the non-random
26:43sampling you are going to applying or
26:45you are going to apply the
26:47the probability sampling but you have to
26:50see you know what exactly the demand
26:52comes from as far as the researcher is
26:54concerned
26:55so the certain time you know what we
26:57observe that when we are doing our
26:59research i think
26:59and we have taken a population or the
27:02universal or we have earmarked the
27:04population we just to narrow down the
27:07research just to
27:08you know cut short the research or try
27:11to
27:11condense the research in a in a shorter
27:14span
27:15we try to follow a fast track mode or
27:17you know the shortcut methods
27:19which is which is no way permissible as
27:22far as the research is concerned so
27:23being a researcher
27:24you have to be quite elaborative you
27:26have to have a patience and you have to
27:28see that how you're going to do the
27:29sampling
27:30because the whole genesis behind the
27:32good research is the sample because
27:34the population is something which you
27:35have earmarked or the universal which
27:37you have
27:38explored this is something which can
27:40give you the boundary
27:41but when you are when you are going to
27:43play in that in that ground or in the
27:45boundary i think
27:46there you have to see because when you
27:49enter into the
27:50into the playground you have to very
27:52easily make up your mind that you're
27:53going to play cricket today you're going
27:55to play football or you're going to play
27:56other games
27:58if you know the intention is not clear
28:01in your mind that what exactly you're
28:02going to
28:03play i think somewhere you are confused
28:05so and uh
28:06it's and that that confusion is going to
28:09eradicate once you
28:10once you know about you know the basic
28:12phenomena of
28:13of the sampling or you know the methods
28:16of sampling
28:17these sampling techniques are are you
28:20know quite vibrant in nature and
28:21it is not like that that one sample you
28:24have used and you are rigid to that
28:26so sometimes what happens we start with
28:27the random sampling
28:29that is stratified sampling simple uh
28:31random sampling or you know
28:33the systematic sampling or you know the
28:35other way of doing the things
28:37but after certain time we realize that
28:39that
28:40that the data is not coming in the
28:41manner we have we have presumers
28:44are supposed to be do so in that case
28:46what exactly you have to do you have to
28:48work as a juggler which i have told in
28:49the in my preceding session also that
28:52you have to do the thing in such a
28:54manner that you can jump to the other
28:56modes of doing the things so if you have
29:00started the journey with
29:01with convenience sampling or the cluster
29:03sampling or the stratified sampling
29:04so because these these are the way by
29:07which you can
29:08able to interpret the particular
29:10clusters or if you talk about
29:12the stratifieds random sampling you can
29:14make certain groups and do the things
29:15but
29:16when you move from this convenience
29:17sampling to judgmental sampling some way
29:19it also helps
29:20because it can give you an uh different
29:22way of doing the things
29:23and on the other hand what we observe
29:25that you have started with the random
29:27sampling then you have
29:28moved to the non random sampling and
29:30apply the snowball sampling now the
29:32snowball sampling
29:33is purely working on the on the
29:34references or to the chain model
29:36so if you if you interact with few
29:39respondents and you ask these
29:41respondents to give
29:42the you know the details about the other
29:45respondents where you can float your
29:46questionnaire i think
29:47this is one of the way by which you know
29:49you can you can get your access more and
29:51and these these respondents considered
29:53to be the potential respondents
29:55it's not like that you can you can
30:00just moving isolately or or
30:04exploding the things in a in a dark room
30:06you are
30:07quite knowing that that you are going to
30:08do the things or you are just you know
30:11going to the right researchers or right
30:12respondents so i think
30:14this is the beauty of the sampling and
30:17and we
30:17we know that when we when we start the
30:19sampling when we do the sampling when we
30:21apply certain techniques or methodology
30:23certain errors comes in between and
30:25these errors work as a alarm
30:27alarm bell and once you get the data and
30:31so the motors are printed like that once
30:33you have reciprocates once once you have
30:35started you know collecting the
30:37uh the data from the respondents in the
30:39primary mode
30:40you try to evaluate those data after
30:43it's not like that
30:44that you have you have a sample size of
30:47of 400 or 500 and 100 people have
30:49reciprocate then you wait for another
30:51400 and then come to the conclusion
30:53it's not like that you the the movement
30:55you know 100 people have reciprocates
30:57you make a small you know competitive
30:59view or
31:00of the of the respondents reports and
31:03try to find out
31:04that what exactly uh you know
31:07the things are coming out if you know
31:09the results considered to be
31:11appreciable or you know the considered
31:14results considered to be
31:17satisfactory then you can carry on with
31:19that otherwise in
31:20in between you have to apply certain
31:22other methodologies which can give you a
31:24model which can give you a platform
31:26that that you reach to a a conclusion so
31:30i think this could be the way and we
31:32have to be quite
31:33quite serious when we are doing
31:37when we are in the process of collecting
31:38a sample because the if you have
31:39collected the sample of 100 respondents
31:42and
31:42and if you have made the competitive
31:43view i think the interpretation could be
31:46quite clear that whether you are moving
31:47a right direction
31:48or you know some some other points need
31:51to be added in questionnaire which can
31:52give you a more
31:53more elaborative result so that is where
31:55you know you have to stop yourself
31:57or you have to revamp your questionnaire
31:59and then you have to start your journey
32:01so
32:01so i think this this research when we
32:04talk about the research when you talk
32:05about the research plan
32:06when we talk about the research design
32:08it's not like that you have make the
32:10research plan or design the very
32:11beginning
32:11you have to be quite you know agile or
32:14or you know
32:16be active as far as the research process
32:18is going on so
32:19every juncture you have to recapitulate
32:21yourself you have to
32:22see whether you are moving in a right
32:24manner or or you know
32:26what certain things would need to be
32:27added or deleted or
32:29you know subtracted so that you can
32:31after certain time you read you need
32:34lead to a final conclusion so anyway i
32:36think we are going to
32:38wind up this particular session which is
32:39quite thought provoking and very
32:41important as far as the
32:42as far as the researcher is concerned
32:44because sampling is is very important
32:46ingredient
32:46you know and in the next session we are
32:49going to talk about the errors they are
32:50margin of errors there are
32:51statistical errors or sampling errors
32:54and how these sampling errors are
32:56are working as a alarm bell and how
32:59you're going to come out because if you
33:00are familiar with
33:01with the with the pros and cons i think
33:04the the result which you are used to get
33:05it's quite foolproof in nature
33:07so thank you learners we will have
33:09another sessions in coming days thank
33:11[Music]
33:22you
33:27you

About this transcript

This page contains the full transcript of Non Random Sampling by MCO-3 [RM&SA], generated from the public captions YouTube serves with the video. The transcript has 6,454 words across 1,133 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.

What you can do with it

Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to translate the transcript into 100+ languages.

Free YouTube transcript tool

YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, translation to 100+ languages, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.