Methodology tutorial - theory-finding research designs: Difference between revisions

The educational technology and digital learning wiki
Jump to navigation Jump to search
m (Text replacement - "<pageby nominor="false" comments="false"/>" to "<!-- <pageby nominor="false" comments="false"/> -->")
 
(4 intermediate revisions by 2 users not shown)
Line 1: Line 1:
{{Incomplete}}
{{Incomplete}}
{{under construction}}


<pageby nominor="false" comments="false"/>
<!-- <pageby nominor="false" comments="false"/> -->


This is part of the [[methodology tutorial]].
This is part of the [[methodology tutorial]].
Line 9: Line 8:
<div class="tut_goals">
<div class="tut_goals">
; Learning goals
; Learning goals
* Learn the names of different kinds of qualitative methdology
* Learn the names of different kinds of qualitative methodology
* Understand the typical research process (which is very different from a theory-testing approach)
* Understand the typical research process (which is very different from a theory-testing approach)
* Be able to differentiate use of qualitative methods in a rather quantiative design and a fully qualitative design
* Be able to differentiate use of qualitative methods in a rather quantitative design and a fully qualitative design
; Prerequisites
; Prerequisites
* [[Methodology tutorial - empirical research principles]]
* [[Methodology tutorial - empirical research principles]]
Line 22: Line 21:
; Quality
; Quality
* Should be ok (but needs to be improved in various ways)
* Should be ok (but needs to be improved in various ways)
* Some translation is needed (examples)
</div>
</div>


Line 30: Line 30:
== The concept of qualitative methodology ==
== The concept of qualitative methodology ==


What is qualitative research and qualitiative methodology ?
What is qualitative research and qualitative methodology ?


* There exist two frequent stereotypes: "Qualitative methodology" is often synonym of "simple description" or "interview analysis". Such a view is held by people who don't understand the difference between a method-as-technique and a method-as-approach.
* There exist two frequent stereotypes: "Qualitative methodology" is often synonym of "simple description" or "interview analysis". Such a view is held by people who don't understand the difference between a method-as-technique and a method-as-approach.
* In reality, there exists a huge pool of design approaches and methods.
* In reality, there exists a huge pool of design approaches and methods.
* Qualitative designs are usually more difficult to create than quantitative designs. Since a typcial humanities student is afraid of statistics, he/she may think that qualitative research is easier, but it's not ...
* Qualitative designs are usually more difficult to create than quantitative designs. Since a typical humanities student is afraid of statistics, he/she may think that qualitative research is easier, but it's not ...


Examples of qualitative approaches (there are more !)
Examples of qualitative approaches (there are more !)
Line 57: Line 57:
| rowspan="1" colspan="1" |participatory design of something
| rowspan="1" colspan="1" |participatory design of something
|-
|-
| rowspan="2" colspan="1" |Design sciences
| rowspan="1" colspan="1" |Design sciences
| rowspan="1" colspan="2" |See [[Methodology tutorial - design-oriented research designs]]
| rowspan="1" colspan="2" |See [[Methodology tutorial - design-oriented research designs]]
|-
|-
Line 87: Line 87:
== The qualitative research process ==
== The qualitative research process ==


Qualitative research usually works in cycles. Its most common features are the following:
Qualitative research is usually carried out in cycles. Its most common features are the following:
* Research must be ''anchored'' in “rich” descriptions
* Research must be ''anchored'' in “rich” descriptions
* Each theoretical ''proposition'' must be anchored in ''observations''
* Each theoretical ''proposition'' must be anchored in ''observations''
Line 93: Line 93:
* Most modern designs also insist on reliability and validity issues.
* Most modern designs also insist on reliability and validity issues.


Regarding the role of theory: 2 very different doctrines:
Regarding the role of theory there exist two very different doctrines. Each has advantages and drawbacks.


{| border="1"
{| border="1"
! rowspan="1" colspan="2" |
! rowspan="1" colspan="2" |Little theory (no grounding of research <br /> questions and analysis grids)
little theory (grounding of research <br /> questions and analysis grids)
! rowspan="1" colspan="2" |A lot of theory (grounding of research questions)
! rowspan="1" colspan="2" |
A lot of theory (grounding)
|-
|-
| rowspan="2" colspan="1" |
| rowspan="2" colspan="1" |[[Image:icon-thumb-up.png]]
[[Image:icon-thumb-up.png]]
| rowspan="2" colspan="1" |openness of mind
| rowspan="2" colspan="1" |
| rowspan="1" colspan="1" |[[Image:icon-thumb-up.png]]
openness of mind
| rowspan="1" colspan="1" |linking to other research
| rowspan="1" colspan="1" |
[[Image:icon-thumb-up.png]]
| rowspan="1" colspan="1" |
linking to other research
|-
|-
| rowspan="1" colspan="1" |
| rowspan="1" colspan="1" |[[Image:icon-thumb-down.png]]
[[Image:icon-thumb-down.png]]
| rowspan="1" colspan="1" |closeness of mind
| rowspan="1" colspan="1" |
closeness of mind
|-
|-
| rowspan="1" colspan="1" |
| rowspan="1" colspan="1" |[[Image:icon-thumb-up.png]]
[[Image:icon-thumb-up.png]]
| rowspan="1" colspan="1" |allows to tackle new subjects
| rowspan="1" colspan="1" |
| rowspan="1" colspan="1" |[[Image:icon-thumb-up.png]]
allows to tackle new subjects
| rowspan="1" colspan="1" |integration of your results with other knowledge
| rowspan="1" colspan="1" |
[[Image:icon-thumb-up.png]]
| rowspan="1" colspan="1" |
integration of your results with other knowledge
|-
|-
| rowspan="1" colspan="1" |
| rowspan="1" colspan="1" |[[Image:icon-thumb-down.png]]
[[Image:icon-thumb-down.png]]
| rowspan="1" colspan="1" |tendency to collect too much data
| rowspan="1" colspan="1" |
| rowspan="1" colspan="1" |[[Image:icon-thumb-down.png]]
tendency to collect too much data
| rowspan="1" colspan="1" |tendency to ignore phenomena
| rowspan="1" colspan="1" |
[[Image:icon-thumb-down.png]]
| rowspan="1" colspan="1" |
tendency to ignore phenomena
|-
|-
| rowspan="1" colspan="1" |
| rowspan="1" colspan="1" |[[Image:icon-thumb-down.png]]
[[Image:icon-thumb-down.png]]
| rowspan="1" colspan="1" |difficult comparison with other work
| rowspan="1" colspan="1" |
| rowspan="1" colspan="1" |[[Image:icon-thumb-up.png]]
difficult comparison with other work
| rowspan="1" colspan="1" |easier generalization
| rowspan="1" colspan="1" |
[[Image:icon-thumb-up.png]]
| rowspan="1" colspan="1" |
easier generalization
|-
|-
| rowspan="1" colspan="1" |
| rowspan="1" colspan="1" |[[Image:icon-thumb-down.png]]
[[Image:icon-thumb-down.png]]
| rowspan="1" colspan="1" |non-explicit preconceptions
| rowspan="1" colspan="1" |
| rowspan="1" colspan="1" |[[Image:icon-thumb-up.png]]
non-explicit preconceptions
| rowspan="1" colspan="1" |explicit preconceptions <br /> (therefore controllable
| rowspan="1" colspan="1" |
[[Image:icon-thumb-up.png]]
| rowspan="1" colspan="1" |
explicit preconceptions <br /> (therefore controllable
|}
|}


=== The description - classification - connection principle ===
Unless you are a student exposed to ethnography, we strongly suggest that you base your research on a '''lot of theory'''. It's safer and simpler...


From Dey (1993:31):
=== Data analysis in structured qualitative research ===
 
In this tutorial, we present a modern, structured view of qualitative data analysis designs
 
Dey (1993:31) formulate the '''description - classification - connection''' principle that you will find in most modern qualitative designs. We are aware that some historical and hermeneutical approaches do not work that way, but for the moment we will not discuss these since they are not very popular in educational technology research.


[[Image:qualitative-research-dey1.png]]
[[Image:qualitative-research-dey1.png]]


[[Image:icon-finger-1-3cm.png|left]] '' description'' : each qualitative analysis relies on “rich” data
[[Image:icon-finger-1-3cm.png|left]] ''description'' : each qualitative analysis relies on “rich” data
:Otherwise you can’t interpret the full meaning of an observation !
:Otherwise you can’t interpret the full meaning of an observation !
<br clear="all"/>
<br clear="all"/>
[[Image:icon-finger-2-3cm.png|left]] '' classification'' : data structuring and
[[Image:icon-finger-2-3cm.png|left]] ''classification'' : data structuring and reduction according to coding principles
reduction according to coding principles
:The mass of data can be staggering !
:The mass of data can be staggering !
<br clear="all"/>
<br clear="all"/>
[[Image:icon-finger-3-3cm.png|left]] '' connection'' : Identification of relationships between concepts
[[Image:icon-finger-3-3cm.png|left]] ''connection'' : Identification of relationships between concepts
:To make relations (and other structure) appear !
:To make relations (and other structures) appear !
<br clear="all"/>
<br clear="all"/>


Here is a dynamic vision of the same principle
Here is a dynamic vision of the same principle, also from Dey (1993:53)
 
Also from Dey (1993:53)


[[Image:qualitative-research-dey2.png]]
[[Image:qualitative-research-dey2.png]]


This figure show the '' circularity'' of a qualitative approach:
This figure show the ''circularity'' of a qualitative approach: After (or even during) classifying and connecting data you will have to look at data again or even produce new data.
 
* classify and connect data
* The need to look at data again or to produce new data


"Modern" qualitative researchers:
"Modern" qualitative researchers:


* '' produce a lot of drawings''
* ''produce a lot of drawings''
* use '' matrices''
* use ''matrices''
* use (sometimes) '' quantitative data exploration techniques''
* use (sometimes) ''quantitative data exploration techniques''
 
Principal difficulty = Do something with the huge mass of data


=== The principle again over time ===
Miles &amp; Huberman (1994:10) illustrate the same principle with a time line and a transition diagram.
Their view is slightly different. "Data reduction" is more or less the equivalent of "classify" and "visualization" the equivalent of "connect".


Miles &amp; Huberman (1994:10)
But the fundamental principle is the same: '''Your principal difficulty is to do something with the huge mass of data in order to find structure'''. In that respect, most qualitative analysis is no different from quantitative analysis.


[[Image:qualitative-research-miles1.png]]
[[Image:qualitative-research-miles1.png]]


A dynamic version of the same schema
A dynamic version of the same schema:


[[Image:qualitative-research-miles2.png]]
[[Image:qualitative-research-miles2.png]]
Line 201: Line 174:


* Qualitative data are most frequently generated by the researcher (same as in quantitative designs)
* Qualitative data are most frequently generated by the researcher (same as in quantitative designs)
* However, qualitative approaches prefer '' natural'' data and refers to the concepts of '' meaning'' and '' process'' (the last issue is shared with systems analysis).
* However, qualitative approaches prefer '' natural'' data and refer to the concepts of ''meaning'' and ''process'' (the last focus is shared with systems analysis and some design research).


Some elements that distinguish between typical quantitative and qualitative research:
Here are some elements that show the distinction between typical quantitative and qualitative research:


{| border="1"
{| border="1"
! rowspan="1" colspan="2" |
! rowspan="1" colspan="2" |Types of approach
Types of approach
|-
|-
! rowspan="1" colspan="1" |
! rowspan="1" colspan="1" |Quantitative approaches search for:
Quantitative approaches search:
! rowspan="1" colspan="1" |Qualitative approaches search for:
! rowspan="1" colspan="1" |
Qualitative approaches search:
|-
|-
| rowspan="1" colspan="1" |
| rowspan="1" colspan="1" |Social or individual structures: ''laws''
social or individual structures: '' laws''
| rowspan="1" colspan="1" |Social construction: ''rules and languages'' as they are''perceived'' and ''created'' by subjects
| rowspan="1" colspan="1" |
social construction: '' rules and languages” '' as they are'' perceived'' and '' created''
by subjects
|-
|-
| rowspan="1" colspan="1" |
| rowspan="1" colspan="1" |Observable facts
observable facts
| rowspan="1" colspan="1" |''Units of meaning'', interpretations by people, e.g. subjective meaning and goals of an action.
| rowspan="1" colspan="1" |
'' units de meaning'' , interpretations by peoplee.g. subjective meaning and goal of an
action
|-
|-
| rowspan="1" colspan="1" |
| rowspan="1" colspan="1" |Abstract behavior and attitudes or experimental situations
abstract behavior and attitudes or experimental situations
| rowspan="1" colspan="1" |Actions and thoughts ''in context''
| rowspan="1" colspan="1" |
actions and thoughts in context
|-
|-
| rowspan="1" colspan="1" |
| rowspan="1" colspan="1" |Standardized macro-observations(applied to a population)
standardized macro-observations(applied to a population)
| rowspan="1" colspan="1" |''Thick'' micro-observations (i.e. few “settings”, small groups, etc.)
| rowspan="1" colspan="1" |
'' "thick" '' micro-observations(few “settings”, small groups, etc.)
|}
|}


== Examples ==
== Examples of qualitative MA thesis ==


* Also consult the module on design-oriented research designs
* Also consult the module on design-oriented research designs
Line 298: Line 258:


[[Category: research methodologies]]
[[Category: research methodologies]]
[[Category: tutorials]]
[[Category:Research methodology tutorials]]

Latest revision as of 18:32, 22 August 2016


This is part of the methodology tutorial.

Introduction

Learning goals
  • Learn the names of different kinds of qualitative methodology
  • Understand the typical research process (which is very different from a theory-testing approach)
  • Be able to differentiate use of qualitative methods in a rather quantitative design and a fully qualitative design
Prerequisites
Moving on
Level and target population
  • Beginners
Quality
  • Should be ok (but needs to be improved in various ways)
  • Some translation is needed (examples)

Most often (but not exclusively) theory-finding research designs use a very qualitative approach whereas theory-testing designs are quantitative. But this is just a general observation, not a necessity. You also could design an exploratory quantitative research design or conduct qualitative research that is strongly theory driven.

Design science most often uses a rather qualitative research approach, but since its purpose is different, we will deal with it in another tutorial (Methodology tutorial - design-oriented research designs).

The concept of qualitative methodology

What is qualitative research and qualitative methodology ?

  • There exist two frequent stereotypes: "Qualitative methodology" is often synonym of "simple description" or "interview analysis". Such a view is held by people who don't understand the difference between a method-as-technique and a method-as-approach.
  • In reality, there exists a huge pool of design approaches and methods.
  • Qualitative designs are usually more difficult to create than quantitative designs. Since a typical humanities student is afraid of statistics, he/she may think that qualitative research is easier, but it's not ...

Examples of qualitative approaches (there are more !)

Family Names Description
investigative journalism case description explanatory story
collaborative research action research practical experimentation with a social goal
participatory observation analytic immersion
collaborative research participatory design of something
Design sciences See Methodology tutorial - design-oriented research designs
language text analysis analysis of relations between elements (grammars)
dialogue analysis organization and structure of dialogue
observation in context anthropology structured and non-structured observation
“field research” (same, but less in depth, more formal)
interpretism hermeneutics human activity as "text"
phenomenology empathy of “Lebenswelt”
symbolic interactionism symbolic interactions between actors

The qualitative research process

Qualitative research is usually carried out in cycles. Its most common features are the following:

  • Research must be anchored in “rich” descriptions
  • Each theoretical proposition must be anchored in observations
  • The researcher plays a delicate role. He (most) always is visible and even can play an active role, i.e. attempt to transform reality.
  • Most modern designs also insist on reliability and validity issues.

Regarding the role of theory there exist two very different doctrines. Each has advantages and drawbacks.

Little theory (no grounding of research
questions and analysis grids)
A lot of theory (grounding of research questions)
Icon-thumb-up.png openness of mind Icon-thumb-up.png linking to other research
Icon-thumb-down.png closeness of mind
Icon-thumb-up.png allows to tackle new subjects Icon-thumb-up.png integration of your results with other knowledge
Icon-thumb-down.png tendency to collect too much data Icon-thumb-down.png tendency to ignore phenomena
Icon-thumb-down.png difficult comparison with other work Icon-thumb-up.png easier generalization
Icon-thumb-down.png non-explicit preconceptions Icon-thumb-up.png explicit preconceptions
(therefore controllable

Unless you are a student exposed to ethnography, we strongly suggest that you base your research on a lot of theory. It's safer and simpler...

Data analysis in structured qualitative research

In this tutorial, we present a modern, structured view of qualitative data analysis designs

Dey (1993:31) formulate the description - classification - connection principle that you will find in most modern qualitative designs. We are aware that some historical and hermeneutical approaches do not work that way, but for the moment we will not discuss these since they are not very popular in educational technology research.

Qualitative-research-dey1.png

Icon-finger-1-3cm.png

description : each qualitative analysis relies on “rich” data

Otherwise you can’t interpret the full meaning of an observation !


Icon-finger-2-3cm.png

classification : data structuring and reduction according to coding principles

The mass of data can be staggering !


Icon-finger-3-3cm.png

connection : Identification of relationships between concepts

To make relations (and other structures) appear !


Here is a dynamic vision of the same principle, also from Dey (1993:53)

Qualitative-research-dey2.png

This figure show the circularity of a qualitative approach: After (or even during) classifying and connecting data you will have to look at data again or even produce new data.

"Modern" qualitative researchers:

  • produce a lot of drawings
  • use matrices
  • use (sometimes) quantitative data exploration techniques

Miles & Huberman (1994:10) illustrate the same principle with a time line and a transition diagram. Their view is slightly different. "Data reduction" is more or less the equivalent of "classify" and "visualization" the equivalent of "connect".

But the fundamental principle is the same: Your principal difficulty is to do something with the huge mass of data in order to find structure. In that respect, most qualitative analysis is no different from quantitative analysis.

Qualitative-research-miles1.png

A dynamic version of the same schema:

Qualitative-research-miles2.png

The role of data

  • Qualitative data are most frequently generated by the researcher (same as in quantitative designs)
  • However, qualitative approaches prefer natural data and refer to the concepts of meaning and process (the last focus is shared with systems analysis and some design research).

Here are some elements that show the distinction between typical quantitative and qualitative research:

Types of approach
Quantitative approaches search for: Qualitative approaches search for:
Social or individual structures: laws Social construction: rules and languages as they areperceived and created by subjects
Observable facts Units of meaning, interpretations by people, e.g. subjective meaning and goals of an action.
Abstract behavior and attitudes or experimental situations Actions and thoughts in context
Standardized macro-observations(applied to a population) Thick micro-observations (i.e. few “settings”, small groups, etc.)

Examples of qualitative MA thesis

  • Also consult the module on design-oriented research designs

... need to be translated some day

Master thesis of M-A Thibaut

Title: Le cartable électronique®. Un Environnement Numérique de Travail en construction. Pratiques éducatives et mutualisation

Research questions (quotes from the thesis: p. 26)

  1. Dans l’utilisation que font les enseignants du cartable électronique®, stabilisent-ils des stratégies pédagogiques ?
    • Recourent-ils à ces outils dans la mesure où cela ne perturbe pas leur habitus d’enseignement ou de gestion de la classe ?
    • Est-ce que l’on voit apparaître la mise en place de scénarios sociopédagogiques, collaboratifs ?
  2. Compte tenu de l’impact du sentiment d’utilité dans l’intégration d’une innovation, pouvons-nous attester dans le cadre de ce dispositif de bénéfices retirés par les enseignants ?
  3. Quelles sont leurs habiletés actuelles en terme de mutualisation au sein du cartable mais également à l’extérieur ?
    • A travers l’idée qu’ils doivent être les constructeurs des contenus pédagogiques du cartable électronique®, quelle est leur position vis-à-vis de cet investissement ?
    • Quelles sont les stratégies qu’ils mettent en place pour exploiter les ressources qu’ils ont à disposition sur Internet et quelles sont les ressources dont ils ont besoin au sein du cartable ?

Method:

(quotes from the thesis: 27-29)

  • Mon travail est une enquête, un regard posé sur les utilisateurs du dispositif. Il est basé sur une étude qualitative et la rencontre de 11 enseignants de différents collèges qui utilisent le cartable électronique® de Savoie. Nous avons préféré questionner plusieurs équipes pour que l’étude soit la plus représentative possible.
  • Le type d’entretien s’oriente vers le « Story telling » ou « récit de vie ». Peu directif, parfois une seule question suffit et ne comporte aucune question directe et fermée. Dans ce cadre, il faut insister sur ce qu’ils font concrètement et essayer de modérer les appréciations sur ce qu’ils pensent. Ils doivent raconter par exemple un événement difficile ou au contraire enrichissant.
  • Mes entretiens ont duré en moyenne 40 minutes. J’ai construit un canevas reprenant mes points principaux. Dans mes entretiens je me suis tenu à rester au maximum sur les faits, les pratiques et les applications mais certains passages renvoient à leurs représentations, passages que j’ai séparés dans mon analyse. C’est la particularité des récits autobiographiques où l’on retrouve toujours deux types d’information : des indications évènementielles (faits, pratiques...) et des réflexions subjectives (représentations, ressentis, avis...)
  • Un questionnaire à la fin de l’entretien leur a été adressé pour connaître leur équipement informatique et leur utilisation à domicile d’Internet. Il se trouve qu’ils sont tous équipés d’une connexion Internet et qu’ils utilisent le web quotidiennement.
  • J’ai rencontré progressivement les enseignants et retranscrit parallèlement les entretiens (Annexe D) ce qui m’a permis de réorienter certaines questions. J’ai traité mes données par rapport à mon canevas (Annexe A et B). Le but de l’analyse est de mettre en évidence les constantes des récits, les régularités mais également les cas de particularité.

Master thesis by D. Touvet

Title: Vers de nouvelles formes d'organisation de l'enseignement. Analyse d'experiences de mediatisation de cours

Research questions (thesis: 11)

Pour chacun des cas étudiés, nous cherchons à savoir :

  1. Comment se déroulent les processus de médiatisation ?
  2. Quelles sont ces nouvelles formes d'organisation de l'enseignement ?
  3. Quels sont les nouveaux rôles qui apparaissent tout au long d'un processus de médiatisation d'un cours ?

Method (quotations from the thesis: 37-48)

  • Nous avons choisi d'effectuer une recherche qualitative selon une approche s'inspirant en grande partie de celle proposée par Huberman et Miles (1991). Ils proposent une méthode de recueil et d'analyse de données qualitatives comprenant les phases de recueil, condensation, présentation et vérification des données.
  • Nous avons décidé de constituer un échantillon de trois projets homogènes, c'est-àdire relativement proches dans leurs caractéristiques pour permettre une meilleure focalisation et comparaison. Ils présentent cependant une bonne variété de choix et de situations possibles.
  • Deux méthodes ont été retenues pour recueillir les données :
    • 1. Passation d'entretiens auprès du coordinateur de chaque projet.
    • 2. Consultation des sites web de chaque projet (deux des trois projets retenus ont noté toute leur démarche sur le site web du projet) ce qui nous a permis d'obtenir des informations pertinentes complétant celles obtenues par les entretiens. Ces adresses nous ont été fournies par les coordinateurs.
  • Le recueil de données s'est effectué à l'aide d'une grille d'entretien dont les différentes rubriques ont été définies en étroite articulation avec la partie théorique de cette recherche. Ainsi, cette grille permet de comprendre de quelle manière se déroulent les processus de médiatisation, quelles sont les nouvelles formes d'organisation de l'enseignement qui se dégagent et quels nouveaux rôles apparaissent tout au long d'un processus de médiatisation d'un cours.
  • Des phases successives de condensation des données recueillies (simplification et synthèse) ont ensuite été initiées et ont abouti à un format de présentation permettant une discussion sur les résultats obtenus.

Bibliography

  • Denzin, Norman K. and Yvonna S. Lincoln (eds) (). The SAGE Handbook of Qualitative Research, Sage Publications. ISBN 0761927573
  • Miles, Matthew B. and Michael Huberman. (1994) Qualitative Data Analysis: An Expanded Sourcebook, Sage Publications. ISBN 0803955405
  • Dey, I. (1993). Qualitative Data Analysis. Routledge, London. * Marshall, C. & Rossman, G. B. (1995) , Designing Qualitative Research, second edition, Sage, London