jamovi

i think there's the theoretical approach, where you're trying to categorize variables precisely (i.e. number of baby spiders is discrete, numerical). and there's a practical approach where we treat it "as though" it's continuous.

i think from a teaching perspective, it's important to hammer home the point that the way we *treat* the variable is sometimes different from the way the variable actually is.

when we perform a linear regression with number of baby spiders as a covariate, it's treated "as though" it's a continuous variable. similarly, we may treat number of human pregnancies as an ordinal or nominal variable.

cheers

jonathon

Every single-word term is clearly wrong. (Ranks are "quantitative", but should be excluded from what's currently labelled "continuous." So "quantitative" is wrong too.)

It's the fault of the entire field of statistical mathematics that there isn't a correct, single-word term.

Here's an invented one: RINTERVAL

Jonathon:

I think this is a distraction from the main issue. Having to call something "continuous" that is clearly discrete doesn't really do anyone a service when setting up data. Mighty confusing, especially when jamovi is promoted on the home page as easy to use and hence good for teaching. And many thanks for you Jonathon and co.

It *is* good for teaching. But my... that is very confusing!

My two cents.

P.

But I see no one has (yet) proposed a correct alternative!

[RE "Having to call something "continuous" that is clearly discrete doesn't really do anyone a service when setting up data."]

just a heads-up, i'm not sure how fruitful subsequent discussion will be.

jonathon

reason180 wrote:But I see no one has (yet) proposed a correct alternative!

[RE "Having to call something "continuous" that is clearly discrete doesn't really do anyone a service when setting up data."]

I proposed "quantitative".

Even if that is not perfect, it surely is less confusing that "continuous".

It simply makes no sense to describe, say, "Number of brothers' as continuous.

P.

But the implication of a non-quantitative "rank" is pretty confusing too, since it's actually false that ranks are non-quantitative.

"Interval/Ratio" is probably the least wrong, most educationally appropriate term. Would you prefer that to "continuous"?

RE: "by kuttu » Sun May 03, 2020 5:05 pm. I'm an undergraduate student considering a career in neuroscience research. So far in all my homework research projects Jamovi is enough. Considering the complexity of SPSS, is it worth learning? In what situations can it specifically help?"

I'm not sure this is the right spot for this comment. but, I don't think SPSS is all that complex. Its file handling may be a little more complex, but jamovi's interface is modeled after SPSS. So I doubt that you would fave difficulty learning to use SPSS.

There are three things jamovi doesn't (yet) do, for which I rely on software other than jamovi.

(1) Adjustable Graphs. It's crucial to able to set the range of the axes for graphs, whether they're used for exploratory or presentation purposes. jamovi does not yet allow such control. SPSS does allow such control (though SPSS's graphing feature are full of bugs (in my opinion), and you would need to learn not just SPSS, but the accumulated folklore for navigating around the bugs).

(2) Data Restructuring. Sometimes you get a repeated-measures data set in "long" format, and you need to convert it to wide format before using jamovi (or SPSS) to conduct a repeated-measures analysis of variance. jamovi does not currently support such conversion, but SPSS does.

(3) Splitting the Data Set. Sometimes you want to repeated and automatically perform the same analysis (such as an ANOVA) on different subsets of a data file. SPSS supports that, but jamovi doesn't.

These are most of the reasons I haven't (yet) totally abandoned SPSS.

just a heads-up,

that post by kuttu is (remarkably sophisticated) spam. this is about the fourth time they've posted that. they normally place a super tiny hyperlink down at the end of the post. i only cottoned on that this was happening a couple of days ago.

jonathon

Oops. Sorry about that (spam).

But, what about "Interval/Ratio"?