Skip to main content
Search form
  • 00:09

    DR. ZINA O'LEARY: If you want to know what you should bedoing--quantitative approaches, qualitative approaches or mixedapproaches.It's good to know where these things come from.So you probably know that quantitative researchis research that really works with numbers and statistics.And it's about representative samples and generalizability

  • 00:30

    DR. ZINA O'LEARY [continued]: and it's large scale survey research,and it's experimental design.And it really does come from a paradigmwhere credibility is seen to come through large samplesand statistical analysis.Qualitative, on the other hand, comesfrom a paradigm where we want to understand depth and richness.

  • 00:53

    DR. ZINA O'LEARY [continued]: And it comes from a world view thatsays we don't have to have massive numbers.We can just look at cases and key informantsand observe almost ideographic thingsand still draw out rich meaning.And then you have mixed approaches.Well, can I just do a bit of both?Are the paradigms too far apart or can I use mixed methods

  • 01:16

    DR. ZINA O'LEARY [continued]: or how do I manage all of that?All right, so for me it really comes down to your question.Sure, there are researchers who are,OK I'm a quantitative researcher, that's what I do.I'm a qualitative researcher, that's what I do.That can be limiting, because it limits the types of questionsyou can ask.I like to start from the premise is, what is your question?

  • 01:38

    DR. ZINA O'LEARY [continued]: Is it a good question?Is it a question of value?Is it a question that can help us make the world a betterplace even in a very small way?So if you get a good research question,then I would suggest the next thing to askis what's the best way to get data to answer that question.I don't premise one type of data over another, one

  • 02:01

    DR. ZINA O'LEARY [continued]: type of paradigm over another.It simply comes down to, what does your question require?How to think about that.Does your question lead you to get responsesfrom a massive group of people?Is your question one that a community needs to answer?If it is, then you probably need to do a survey.And a survey will require statistical analysis,

  • 02:22

    DR. ZINA O'LEARY [continued]: and you're in a quantitative paradigm,collecting quantitative data.But what if your question is one that'sbest answered by talking to an individual who has experiencedthis, who's felt it, whose felt whatever the emotional responseis to it.

  • 02:43

    DR. ZINA O'LEARY [continued]: So if you want to know how a community responds to a naturaldisaster, you might survey the community--quantitative.If you want to know how an individual respondsto losing their home and their pet,then you're in the qualitative realm, when you reallyneed to talk to them.If you want to understand the whole experience,

  • 03:03

    DR. ZINA O'LEARY [continued]: then you might use the mixed approachwhere you're doing a bit of each.So the answer to quant, qual or mix is that none of themare better than the other.They are better only to the extentthat they help you answer your research question.And that is paramount.So once you realize who has the answer to your question,

  • 03:24

    DR. ZINA O'LEARY [continued]: then figuring out how to best getthe answers from that group, will determine your method.But you also need to think about what data you actuallyhave access to.Can you reach the people you want to reach?If you have key informants or particular cases,will you have access?Surveying is not that easy either,

  • 03:45

    DR. ZINA O'LEARY [continued]: because we have to have a good response rate.Let's say you can send out survey nowon the Internet to 200 people.And you get a 10% response rate so you have 200 surveys back.And you think that's not too bad, with 200I can do all my statistical analysis.I'm fine.Well you have to ask yourself, is the 200 who answered,

  • 04:06

    DR. ZINA O'LEARY [continued]: are they different, significantly different,in any way, to the 800 who didn't bother.All right?So you'll make false assumptions aroundgeneralizability if you have what'scalled a non-response bias.So do you have access to data?Think about how much time you have.Do you have the time and the skills

  • 04:27

    DR. ZINA O'LEARY [continued]: to do qualitative, in-depth research?Do you have the time and the skills to do a survey?Do you have the time to try to do two different thingsand have that mixed approach?So the challenge is to think about,do I want to do the quant and how will I do it?Do I want to do the qual, and how would I do it?Or do I look at the mixed methods?

  • 04:48

    DR. ZINA O'LEARY [continued]: Now mixed methods are becoming ever more popular.There's something that people think, all right,well there's limitations in quantitative data,there's limitations in qualitative,let's have a mixed approach.Let's do some document analysis, some key informant interviewsand do a survey and we'll get the full picture.And that's fine, given you have the time and the skillsto do so.

  • 05:09

    DR. ZINA O'LEARY [continued]: So mixed methods allows you to really draw on both traditions,and there's a number of ways you can do it.You can really have a strongly quantitative approach,but kind of add qualitative color.So you do a survey and then you find out sort of the answersand where you're leaning.And then you might pick a couple of peoplewho are identified in that as people you

  • 05:31

    DR. ZINA O'LEARY [continued]: want to talk to and go deeper.And you kind of give some anecdote to your survey.You can do it the other way around.You can use some quantitative data to give weightto qualitative narrative.So you have your rich story that'scoming from the in-depth interviews,almost like a phenomenological approachto understanding what's happening to people.

  • 05:51

    DR. ZINA O'LEARY [continued]: And then you might also have some quantitative statsto give it some credit.You could also do it as equally weighted partners.All right, so you're working with quantitative data,you're working with qualitative dataequally, and analyzing them as though theyhave the same weight.Again, ask me what's the best approach?

  • 06:13

    DR. ZINA O'LEARY [continued]: I'm going to come up with the same answer.The best approach is the one thatgets you credible data that answers your question.That's always the criteria.That's the number one criteria.The other things are secondary.So does the approach best answer your questionin a way that's going to have credibility?And that's because all of these methods have their challenges.

  • 06:38

    DR. ZINA O'LEARY [continued]: All right?If you do quantitative, you reallyhave to guard against the critique that it's too thin.That you've only scratched the surface.That you're only getting the answers to the questionsyou thought to ask.That people don't really--they may just check the same thingall the time in a survey.You don't really know what they're thinking.

  • 06:60

    DR. ZINA O'LEARY [continued]: They're giving practiced responses.Those are the challenges of quantitative data.You're not getting the right sample.The statistics can be manipulated.So lots of challenges with quantitative datathat you're going to have to think of.Equally, there are challenges.If you use a qualitative approach, one of the critiquesis that it's too anecdotal.

  • 07:21

    DR. ZINA O'LEARY [continued]: It's a case study.It's not representative.It's not generalizable.How are you going to have credibility when you onlytalk to three people?How are you going to have credibilitywhen your sample does not representthe general population?And you have to be ready with those arguments.And you have to know why your qualitative approach best

  • 07:42

    DR. ZINA O'LEARY [continued]: answered the question.For me a lot of this comes up with the type of question.If you want to know the demographics of the situation,you probably have to do something quantitative.But if you want to know how something happens, whatthe experience of something is, then it'sprobably going to lend itself to a qualitative approach.So each method inherently has its own challenges.

  • 08:06

    DR. ZINA O'LEARY [continued]: There's no doubt about that.But we often see a mismatch as well.So they have their own challenges.They also have their own strengths.Absolutely, can be creditable methods if well handled.But one of the big challenges is when there's a mismatch.So you have a quantitative-type question

  • 08:27

    DR. ZINA O'LEARY [continued]: where you get qualitative data, or youhave a qualitative-type question and youtry to smash it with numbers.It's not going to work.The other mismatch happens when youdo have the qualitative data and you don't know how to treat it.And you end up doing something so pseudo-statistical with it.So you have seven people and so that'llbe 40% of people think this.

  • 08:48

    DR. ZINA O'LEARY [continued]: Forty percent of seven?Why do I need to know that?It almost assumes that it means 40% of the world,but of course, it doesn't.Because you're seven isn't going to represent anythingin particular.So you have to have the right question whereseven people can illuminate.And you also have to treat the data welland know how to analyze it.The biggest critique of qualitative research

  • 09:11

    DR. ZINA O'LEARY [continued]: is that it's not credible.And the only way to fight that critiqueis by handling it appropriately.So those are the challenges, and of course,with mixed approaches, you have challenges as well.And that it's too resource intensive, would be a main one.Do you have the time and the energyto do both sets of data collection?

  • 09:32

    DR. ZINA O'LEARY [continued]: The other challenge is how do you integrate it?Do you talk about the qualitativeand then the quantitative?Do you treat them equally?How do you mix the paradigms if theyhave different assumptions?So lots of things you have to work through and for allof these traditions, what we're talking about is credibility.In order to get that creditability,we need to make sure whatever approach you

  • 09:54

    DR. ZINA O'LEARY [continued]: pick is the one that best answers the question.And that you are using those methodswith rigor and with a design thatensures that things are being treated such that credibilitybecomes assured.So it really is a management process.So what is best?

  • 10:14

    DR. ZINA O'LEARY [continued]: What's best is what will get you a credible answerto the question with integrity.

Video Info

Series Name: Doing Your Research Project

Episode: 4

Publisher: SAGE Publications Ltd

Publication Year: 2017

Video Type:Tutorial

Methods: Mixed methods, Quantitative data collection, Qualitative data collection

Keywords: challenges, issues, and controversies; criticism and critics; resource management; Skills (abilities); time factors

Segment Info

Segment Num.: 1

Persons Discussed:

Events Discussed:



Dr Zina O'Leary discusses whether qualitative, quantitative, or mixed methods are best. Each approach has its benefits and drawbacks. The best approach is whichever method answers the research question with credibility.

Looks like you do not have access to this content.

What Is Best: Quantitative, Qualitative Or Mixed Approaches?

Dr Zina O'Leary discusses whether qualitative, quantitative, or mixed methods are best. Each approach has its benefits and drawbacks. The best approach is whichever method answers the research question with credibility.

Copy and paste the following HTML into your website