COMParative methods for the Advancement of Systematic cross-case analysis and Small-n Studies

International ressource site

Welcome

Who are we?

COMPASSS activities

NEWS: training, meetings, publications

Bibliographical database

Software

Didactics: Boolean algebra and QCA logic

Working papers

SmallN people

Useful links

CCM Textbook

Didactics (1)

(Francais ici)

Benoît Rihoux (UCL - COMPASSS), Gisèle De Meur (ULB - COMPASSS), Sakura Yamasaki (UCL - COMPASSS) and Sophie Ronsse (UCL - COMPASSS)

1. Some introductory remarks (1)

QCA is still a young method. That is why it is very easy to misunderstand some of its logic or functions, and even sometimes to really misuse it. There is thus a spontaneous need to produce some indicators in terms of good practices for QCA analyses. This document, which is the result of our own experience both as researchers and professors, aims at making an inventory of these good practices (2).

If you wish to add any ideas to this inventory, or any precision or nuance on such and such point, please don't hesitate at letting us know by sending an e-mail to G. De Meur (gdemeur@ulb.ac.be ). Your pertinent suggestions will enrich this document (reference to your name will be made, of course).

This document contains two sections : first, it gives general advice, followed by a second section which provides more specific advice, listed in the sequence of a classical QCA analysis. We hope this will serve as a " check list " for those engaged in a QCA analysis.

2. General advice :

1. Make a reasonable usage of QCA, which can be used in different ways (Click here to have a list of possible functions);

2. Draw on the different functions of the software (many of which are under-used, like the " hypothesis testing " function (3)) ;

3. Use technical and reference concepts with precision, in order not to induce the reader in error (for example, a " condition " is not an " independent variable ", etc.) ;

4. Never forget the configurational logic of QCA : never consider the influence of such or such condition in an isolated manner, especially in the interpretation of minimal formulae ;

5. Never use QCA based on a " press button " logic, but as a tool which requires iterative steps, different go and back towards cases and theories ; do not try to hide the encountered difficulties but explain the way in which they have been solved ;

6. Do not invent what QCA does not tell you. In particular, be careful before interpreting a minimal formula in terms of causality. The formulae express, more modestly, co-occurrences.

7. In the research process, do not exclusively focus on QCA (nor on one and only one method, whatever it may be). At different stages of the research, choose the best fit method according to the needs, the situation, the type of data, or the pursued aim, etc. Use QCA in some stages of the research, et draw lessons and conclusion from other methods at other stages of the research, be these methods qualitative (4) or quantitative (5).

3. Specific advice :

Following the typical QCA application process :

1. Choose the cases in a rigorous manner (truly under a comparative design) ;

2. Develop an " intimacy " with each of the cases ;

3. Choose the conditions in a rigorous fashion, i.e. based on theoretical and empirical grounds, and in reduced number (cf. the ration between the number of cases and the number of conditions) ;

4. When the raw data (not yet dichotomised) are quantitative data, show them ;

5. Clearly mention the way in which each condition is dichotomised. Justify the threshold based on real knowledge and not on pure technical criteria;

6. Show the truth table, by indicating to which observed cases correspond each configuration ;

7. If the truth table shows contradictory configurations, resolve them. There are several ways to do this, cf. Ragin (1987), pp.113-8;

8. Proceed systematically to four minimisations : those for the configurations with an " 1 " outcome, first without and then with the inclusion of logical cases, and those for the configurations with a " 0 " outcome, first without and then with the inclusion of logical cases ;

9. The minimisations should be done with the software and not by hand ("exact" algorithm of Quine-McCluskey);

10. Resolve the potential " contradictory simplifying assumptions " (or " contradictory logical cases ") produced along with the minimisations with inclusion of logical cases (6);

11. Present (even in a summary manner) the main iterations of the research (back to cases, back to theories, model refining, etc…) ;

12. For each minimisation, present all minimal formulae obtained at the end of the analysis. If one of them is eventually selected as best, mention the justification for such a choice;

13. Proceed to a real return to the cases (and/or theory) at the end of the analysis, based on the minimal formulae.

(1) Although QCA is both an approach and a technique, these good practices essentially revolve around technical matters. To have an overview of the on-going discussion around QCA as an approach, see RIHOUX B. (forthcoming), "Bridging the gap between the qualitative and the quantitative worlds? A retrospective view on Qualitative Comparative Analysis", in Field Methods, Vol. 15 (2), pp.115-130.
(Go back to text)

(2) In fact, shouldn't any technique (we are thinking in particular of statistical techniques) equally be accompanied by such a list, a " code of good conduct " which would be a reference in the scientific community ?
(Go back to text)

(3) See WATANABE T., "Where theory and reality meet: using the full potential of QCA by exploiting the intersection function of the QCA software. International comparison analysis about the occurrence of social movement.", Communication presented at the COMPASSS Launching Conference, 16-17 Sept. 2003, Louvain-la-Neuve and Leuven, Belgium. Can be downloaded on the COMPASSS site, "Working Papers" section : http://www.compasss.org/WP.htm ; see also YAMASAKI S., "Testing hypotheses with QCA: application to the nuclear phase-out policy in 9 OECD countries", Communication presented at the 2nd ECPR General Conference, 18-21 Sept. 2003, Marburg, Germany.
(Go back to text)

(4) A good example : BELL J. and LOANE S., "Entrepreneurship research in Europe: innovative methods in the exploration of internationalisation issues", 2003, project in progress, downloadable at www.epi-entrepreneurship.com/doc/BELL.pdf.
(Go back to text)

(5) A good example : BOSWELL T. and BROWN C., "The scope of general theory. Methods for linking deductive and inductive comparative history", Sociological Methods and Research, Vol.28, n°2:154-185, 1999.
(Go back to text)

(6) About these contradictory logical cases, see RIHOUX B., Les partis politiques : organisations en changement. Le test des écologistes, Paris, L'Harmattan, (2001) pp.205-8.
See also VANDERBORGHT Y. and YAMASAKI S., "The problem of contradictory simplifying assumptions in Qualitative Comparative Analysis (QCA)", paper presented at the 2nd ECPR General Conference, 18-21 Sept. 2003, Marburg, Germany. Downloadable at : http://www.essex.ac.uk/ecpr/events/generalconference/papers/6/5/Vanderborght.pdf .
See also VANDERBORGHT Y. and YAMASAKI S., "Des cas logiques… contradictoires ? Un piège de l'AQQC résolu à travers l'étude de la faisabilité politique de l'Allocation Universelle" in Revue Internationale de Politique Comparée, Vol.11 (1), 2004.
(Go back to text)

"QCA lends itself to several functions: it allows either to present the data in a more parsimonious manner, or to validate or refute models or existing theories, or yet to elaborate new theoretical proposals in the form of hypotheses"(p.35; our translation). For more details on each of these functions, see pp.78-80 of the book (in French):

G. DE MEUR et B. RIHOUX, "L'Analyse Quali-Quantitative Comparée (AQQC-QCA). Approche, techniques et applications en sciences humaines", Louvain-la-neuve: Academia-Bruylant, 2002, p.35. Click here to go on the information page about this book.

(Go back to text)

Go back to top

Last modified: 21-Déc-2006

Go to top

Go to top

Go to top

Go to top

Go to top

Benoît RIHOUX, Centre de Politique Comparée