Anca Florescu-Mitchell May 13, 2014:

@ DLyons This is psychology now :)! You are right.

DLyons May 13, 2014:

@Anca This discussion entry will be deleted!

Why not? Because I filtered out this asker a few days ago - there are some people whom it is just too difficult to help.

Anca Florescu-Mitchell May 13, 2014:

@ DLyons Why not? You were the only one who firmly believed it was not about statistics :).

DLyons May 13, 2014:

@all Why not post an updated answer with what you now understand by the term? I'd be happy to give it retrospective support, just to try to set the record straight. I won't be posting anything myself :-)

Anca Florescu-Mitchell May 13, 2014:

@ DLyons But nobody says it is about statistics... anymore ::)). Since you removed your answer, how can we now agree with you?

DLyons May 13, 2014:

At the risk of repeating myself, I don't believe that this sentence actually relates to Statistics. IMHO, it's using a rather modish polysemous French word to mean something akin to "is multidimensional / involves multiple resources/trajectories".

Anca Florescu-Mitchell May 13, 2014:

@ B D Yes, this was the problem all along: to be or not to be... in the context of a multimodal distribution. I have state-of-the-art knowledge about statistics (long story:), but no experience in sociology, so I was really curious to understand what it was about. You are perfectly right, the clarification should have come from the French camp ::))!

B D Finch May 13, 2014:

@Anca To clarify my comments below, I now think that "modalité" is not about mode/multimodal distribution etc. I also believe that it is possible that some statistics terminology may have changed since I studied the subject. The ISI glossary entry suggests to me that some standardisation between languages may have occurred. Perhaps somebody with more up-to-date knowledge of the field than I have needs to look at the French definitions and confirm whether the term "modalities" (nothing to do with "modes") would nowadays be used in English for the same concept.

Anca Florescu-Mitchell May 12, 2014:

@ B D Finch This is why it was important to get a straight answer about the use of the multimodal approach. I have in front of me the Oxford Dictionary of statistics with its article about MODE - related to modal frequency. There is no MODALITY as a statistical term. The internet dictionary you are mentioning is not entirely reliable. Sometimes applied sciences distort the original term. So maybe, if sociologists call a "mode" (same word in French and English) "modalité", you can use "modality" in English. From my perspective though, this is not correct.

B D Finch May 12, 2014:

ISI - INTERNATIONAL STATISTICAL INSTITUTE Glossary http://isi.cbs.nl/glossary/term2129.htm Translates FR Modalité as EN modality, so you may have guessed correctly.

B D Finch May 12, 2014:

@Asker Anca's comment below made me think about whether I had misunderstood the French term "modalité". Looking at the refs. below, I think that "modalité" is actually what one would probably call in English "condition", even though I have now found a glossary that translates it as "state". However, I am now rather less sure about how to translate the term in this context and suggest that it might be a good idea to re-open the question.

"dans une distribution statistique, le mode est la modalité de la variable à laquelle est associé le plus grand effectif ou la plus grande fréquence. On note généralement le mode : M0."
http://www.maxicours.com/soutien-scolaire/ses/2de/223351.htm...

"Un caractère étudié peut être soit qualitatif, soit quantitatif (appelé aussi variable). On parlera alors de modalités d'un caractère qualitatif et des valeurs d'une variable."
http://fr.wikibooks.org/wiki/Statistiques/Généralités

This has made me get an old Statistics textbook off the shelf and realise how long it is since I last referred to it! (And how old it is.)

Anca Florescu-Mitchell May 12, 2014:

@ Asker The outcome is really upsetting.
If you decided this is about statistics after all, look at least at very general presentations such as
http://en.wikipedia.org/wiki/Bimodal_distribution
http://en.wikipedia.org/wiki/Mode_(statistics)
Multi-modal distributions involve modes, not modalities.

B D Finch May 12, 2014:

@Donal You are right in thinking that the statistics involved is very simple. Basic stuff that one would learn in the first year of degree study (before having the opportunity to drop stats from your course units ;)).

DLyons May 12, 2014:

@Barbara I don't disagree. My own gut feeling is that this group hasn't yet got beyond simple descriptive Statistics but, not having access to the full paper, only zac can say :-) //And he or she still isn't saying, despite being asked several times. I rather regret having researched these two questions, but it's not a mistake I'll make again :-)

B D Finch May 12, 2014:

@Donal My degree in Sociology included course units on statistics in all three years and I have some experience of teaching Sociology. So, please take my word for it that Sociology uses statistics and that sociologists tend to use and to be careful not to misuse, statistical terminology.

Anca Florescu-Mitchell May 12, 2014:

@ DLyons Exactly :). I would really like to know!

DLyons May 12, 2014:

@zac I originally thought the term really wasn't a Statistics one, the group's terminology seemed closer to Sociology. Now, I just don't know.

Points are not the issue :-)

Anca Florescu-Mitchell May 12, 2014:

@ Asker What about the actual question, about statistical methods being used?

Zonia Clissold (asker) May 12, 2014:

To DLyons
You are quite right to be irritated by "hit and run" askers. I tend to leave asking questions here to the very last minute with the deadline over my head so don't always reply to queries but make a decision on the first replies I received. I did check out whether any similar surveys had been conducted but did not come across the one you indicated which was extremely helpful. Why don't you put it in answer rather than a discussion then I can give you points?

B D Finch May 8, 2014:

@Donal You have used the researchers' terminology, but applied it to a different context. Social scientists use both statistical and descriptive methods and there is little point doing the statistics if you don't then discuss and draw conclusions from the results, which involves the description. There is also little point collecting data if one does not apply statistical analysis to it, except for cases where the conclusions from the data are so blindingly obvious that you almost wonder why any research was necessary. In the text posted by the Asker, "Mesures et capture" and " la distribution présente trois ou davantage de modalités" both very strongly suggest statistics to me.

Nikki Scott-Despaigne May 8, 2014:

A number of teams are working in this area. The poster in your reference includes data which are descriptive (and which may have been drawn from a body of inferential statistical data). In any event, in the social sciences, the statistical terminology is every bit as precise, in both descriptive and inferential statistics. (Had this hammered home to me in my biology and neuroscience research masters and am seeing it again in the psychology degree I'm doing now). Matters of distribution are very much part of descriptive stats too.
Further context would not go amiss though! It never does. ;-)

Anca Florescu-Mitchell May 8, 2014:

@ Asker Could you please tell us if there is any statistical method involved in this survey?