Skip to main content
Search form
  • 00:00

    [MUSIC PLAYING][SAGE video][An Introduction to Interaction Analysis][Dr. Carol Rivas Senior Research Fellow at Universityof Southampton]

  • 00:10

    CAROL RIVAS: Hello, my name is those Carol Rivas.I'm a senior researcher at the University of Southampton.And I have done a number of studiesusing interaction analysis.And today, I'm going to talk about the different typesof interaction analysis that you can do.Modern day interaction analysis began in the 1960s with someone

  • 00:32

    CAROL RIVAS [continued]: called Garfinkel, a sociologist whodeveloped a branch of qualitative analysis calledethno-methodology.Ethno means people, so ethno-methodologyis the study of people.But what Garfinkel was interested inwas how people interacted with their environmentsin their everyday lives.So what he looked at was the study of the mundane, studies

  • 00:55

    CAROL RIVAS [continued]: of people talking at the dinner tablepassing each other in the street, and so on.And one of the early experiments that Garfinkel didwas to send his students home and pretend to their familiesthat they didn't know their families and see what happens.

  • 01:15

    CAROL RIVAS [continued]: And this is called a breaching experiment,because you breach social norms.So the students went home, and their parentsmight say things to them like, how was your day.And the students would reply something like, well, what's itto you, you don't really know me?And the students reported back at the end of the weekendto Garfinkel that parents had got really upset

  • 01:36

    CAROL RIVAS [continued]: and though that their child was ill and maybe needed treatment.So this was an illustration of how, if you break social norms,you can have problems and be labeled as ill or deviantin some way.At the same time as Garfinkel was doing his experimentation,Erving Goffman, another sociologistwas making more theoretical and observational analyses

  • 02:00

    CAROL RIVAS [continued]: wrote a book The Interaction Order.And in this, he said, it's really vitalthat you keep the conversation or the communicationor interaction going.As soon as interaction breaks down, there's a social problem.And what you have to do in that case is to repair the problem,so you're not labeled as deviant.And so Goffman said that people need

  • 02:23

    CAROL RIVAS [continued]: to do that by making excuses for their behavioror apologizing for it or justifying it in some way.[Conversation Analysis]So a group of students, some of whomwere followers or students of Garfinkel or Goffman,got together and developed a branch of ethno-methodology

  • 02:44

    CAROL RIVAS [continued]: called conversation analysis.So this is the second type of interaction analysisI want to consider.And so one of the key points of ethno-methodologywhich they took across into conversation analysisis that the people in the interactionare interpreting the interaction.The analyst is simply an observer on the outside.

  • 03:08

    CAROL RIVAS [continued]: So for example, going back to this idea of a meetingsomeone in the street.If I meet you in the street, I don't know you very well,and I say, how are you?You're expected to simply say, fine, thanks, and how are you?So we're interpreting the fact that wedon't know each other very well, and actually we'renot interested in how we are.If I meet you, and I say how are you, and you say,

  • 03:30

    CAROL RIVAS [continued]: oh, I'm having a dreadful time of it for recently.And then I say, well, let's go for a coffee,we're intercepting it together as the factthat we know each other well enough to explore this further.And this also is important to conversation analysis,and in particular, you can see from that example,that actions are created through conversation.

  • 03:52

    CAROL RIVAS [continued]: So if we don't get involved in a detailed discussion about howwe are, then we just carry on walking.So the action that's created there is to not do very much.But if we go and have a coffee, a different actionis created through the talk.So this is essential to conversation analysis.Conversation analysis says that the whole context

  • 04:13

    CAROL RIVAS [continued]: is in the talk.You can determine the context in the talk, whichmakes it very different to other types of qualitative research,in which the context is said to exist externally only.Conversation analysis doesn't allowyou to consider the external context,but the focus is on the context in the talk.And what do I mean by external context?

  • 04:35

    CAROL RIVAS [continued]: I mean things like the setting, the room or the streetthat you're in and the time that the interaction istaking place.So fundamental to conversation analysisis the fact that nothing in talk is redundant, disorderly,accidental.And so conversation analysts willanalyze every single bit of talk from the tiniest micropause.

  • 05:00

    CAROL RIVAS [continued]: Because of the detail that conversation analysis go into,they usually only consider one small part of an interaction.So for example, there are quite numberof studies on how people start a conversation offor how they end a conversation.And what they look for in the conversation

  • 05:21

    CAROL RIVAS [continued]: is the types of words that people use,not for the meaning that they have but to the waythey could be used strategically.So if you say, I'm just coming, the justmeans that, really, you shouldn't be so impatient.I am coming.So the little ways that words can be used to modify the way

  • 05:43

    CAROL RIVAS [continued]: you're trying to communicate something.And conversation analysts will alsolook at the rhythm of the talk, the prosody.So whether your voice raises and lowers.And they will look at the speed that youtalk, at the quietness or the loudness that you talk.Also very, very important in conversation

  • 06:04

    CAROL RIVAS [continued]: analysis is to consider the talk that was justbefore what you say and the talk that was just after,because you're making meaning through the talk.And you're interpreting, through the talk,and other people are interpreting the contextthrough your talk.And they will feed back to you, their interpretationthrough their response, meaning that how you progress the talk

  • 06:28

    CAROL RIVAS [continued]: will be affected by people's responses.And because this is so important,you should also consider where the little bit of conversationsits within the larger conversation.So conversation analysis is very much a sequential analysis.To give you some very simple examplesof what I mean by how you use talk,

  • 06:51

    CAROL RIVAS [continued]: if you consider that I'm phoning my boss,and he answers the phone.And I say, I've crashed my car.We are both likely to understand from the intonation of thatthat and I'm going to say, so I can't comein today or something similar.Whereas, if I phone and say, I've crashed my car,

  • 07:12

    CAROL RIVAS [continued]: my boss is likely to think that I need help in some way,because there's some element of panic in the waythat I've said exactly the same words.So that's how conversation analysiscan pick up on context.So now I'm going to try and show youmore of the power of conversation analysis.These excerpts that I'm going to show

  • 07:32

    CAROL RIVAS [continued]: you come from a study I've undertaken, in which I audiorecorded the stop smoking consultations in communitypharmacies in London.So the advisor is A in the excerpts.And the smoker is S. This is a second or third consultationin the series.

  • 07:54

    CAROL RIVAS [continued]: And at the first consultation, the smokeragrees to stop smoking.So by the time of this consultationthat I'm showing you, they should've actually stopped.So in this consultation, you can seethat the advisor is asking the smoker how many cigarettesthey're still smoking, and is asking for some good news,

  • 08:14

    CAROL RIVAS [continued]: hinting that they want to hear that the smoker stopped.So this is a simple typist's analysis.It is not conversation analysis.And from this, we can see that the smoker saysthat they've come down, six, seven, six, they've come down.And so the adviser says, oh, that's really fantastic.

  • 08:36

    CAROL RIVAS [continued]: They get really excited about this.If you do the maths here, they suggestthat when the smoker came to stop smoking,they were smoking 10, because the advisor saysyou're only smoking 25% of what you were smoking before.And they've come down seven.And so from this, the smoker wasn't smoking that manyto start with, you can guess.

  • 08:59

    CAROL RIVAS [continued]: And the advisor is getting, perhaps,a bit excited about this.If we then look at an extract whichhas had a annotation used in conversation analysis, whichis an annotation developed by Gail Jefferson, who I mentionedbefore, an internationally understood annotation,we can see something rather different is happening.

  • 09:21

    CAROL RIVAS [continued]: So in this annotation, the little dots and bracketsmean that there's a micropause, very, very tiny pause.The square brackets show that there's overlapping talk.The numbers in brackets in 0.1 and 0.2,mean significant pauses.0.1 seconds might not sound significant,

  • 09:41

    CAROL RIVAS [continued]: but conversationally, it is significant.And up arrows mean the person is raising their voice.Down arrows, they're lowering it.And the tiny little degree symbolsmean that the person is talking quietly.So if you see exactly the same words with this annotation,what you can see is the smoker says seven,

  • 10:04

    CAROL RIVAS [continued]: and the advisor immediately latches onto thisand thinks that they are only smoking seven now.And says they have come down.So their "whoa" overlaps the rest of the smoker's talk,and they don't hear that they've come down by seven,they think they're only smoking seven.And this is why the advisor gets quite excited,because in fact, if you do the maths,

  • 10:26

    CAROL RIVAS [continued]: you will find that the smoker wassmoking 25 to 30 cigarettes when they came in to stop smoking.And we actually know this, because we collected datafrom the smoker at the start.And so to come down from 25 to 30 to only sevenwould be an achievement.But what's actually happened is the smokeris still smoking around 20.

  • 10:47

    CAROL RIVAS [continued]: And the advisor is not picking this up,because they've talked over the smoker.So you can get this richness from the conversation analysis.And what you can also see is the smokersays "yeah" very quietly, because they don't reallyunderstand why the advisor's getting so excited,because they haven't done very well.They were meant to stop smoking not stillbe smoking 20 cigarettes a day.

  • 11:09

    CAROL RIVAS [continued]: And so the advisor carries on trying to motivate themand can't understand why the smokers isn't gettinghappy about their success and getsreally extreme in trying to push the smoker to gets excited.And when they say they're lost for words,the advisor is truly lost for the words.You can see from the conversation analysisannotation that they're stuttering a lot.

  • 11:30

    CAROL RIVAS [continued]: And you could also see at the end of this extractthat the smoker really is struggling to say anything,because they know they haven't done well.So you can see from this example that youcan get much, much richer and more accurateinformation from conversation analysisthan from a simple typist's transcript, in which you mightmisunderstand the situation.[Multimodality Analysis]

  • 11:54

    CAROL RIVAS [continued]: The third type of interaction analysis you can dois known as mulitmodality analysis.It's fairly recent, and it tends to focuson the interaction between people and computersand other bits of technology.And the experts in this area are based at the Instituteof Education in London.

  • 12:14

    CAROL RIVAS [continued]: And they have done studies that don't just consider technology.For example, they have some researchlooking at surgeons and the way they and their assistantshandle the surgical equipment during surgery.[Conclusion]To conclude, I'd just let's go through the basic steps

  • 12:35

    CAROL RIVAS [continued]: in doing an interaction analysis.The first thing you need to do is collect the data.And because you might only focus on, say, the opening sequence,you need to collect a lot of examples of this.You then need to be able to log, in some way, wherethe interesting bits of the interaction that's takingplace.And so you either need to write these down

  • 12:57

    CAROL RIVAS [continued]: on a typewritten transcript.Or if it's a video, you need to keep a careful log.The next thing you need to do is to code your data.So if it's a conversation analysis,you would use Jeffersonian annotation on a transcript.For video analyzes, multimodality analyzes,

  • 13:18

    CAROL RIVAS [continued]: there is no consensus on how to do it.And everyone at the moment tends to adapt their own method.There's a more general type of codingthat you can do to get the bigger picture,and this is reserved at the moment for medical encounters.The technique that's been developed for thisis called the RIAS, Roter Interaction Analysis System.

  • 13:42

    CAROL RIVAS [continued]: And this will look at things like the amountof biomedical talk that is done in the conversationor the number of questions that are asked.Once you've done your coding, it'sa good idea to hold data workshops,because in interaction analysis, youare the outsider, the observer, looking in other people'sinterpretations.

  • 14:03

    CAROL RIVAS [continued]: You need to make sure you have appreciatedtheir interpretations.So in the data workshops, you tendto play back points of interest in the interaction,and get other people to comment on howthey see the interactants have interpreted things.So from this video, I hope you'vegot a feel for the different types of interaction analysis

  • 14:25

    CAROL RIVAS [continued]: that you can do and the types of datathat you can obtain from this, and alsoget an understanding of the stepsthat you need to go through in such analyses.[SAGE video]

Video Info

Publisher: SAGE Publications Ltd

Publication Year: 2017

Video Type:Tutorial

Methods: Interaction order, Conversation analysis

Keywords: practices, strategies, and tools; Social norms; Social norms and conformity; technology

Segment Info

Segment Num.: 1

Persons Discussed:

Events Discussed:



Dr. Carol Rivas discussesthe different types of interaction analysis. According to interaction analysis theory, when interaction breaks down, it causes a social problem and the problem has to be repaired so it does not become deviant. Rivas discusses interaction analysis, conversation analysis, and multimodality analysis.

Looks like you do not have access to this content.

An Introduction to Interaction Analysis

Dr. Carol Rivas discussesthe different types of interaction analysis. According to interaction analysis theory, when interaction breaks down, it causes a social problem and the problem has to be repaired so it does not become deviant. Rivas discusses interaction analysis, conversation analysis, and multimodality analysis.