Wikidata:Property proposal/Data analysis method
Data analysis method
editReturn to Wikidata:Property proposal/Generic
Under discussion
| Description | methods used in the main item for inspecting, cleansing, transforming, and modeling data with the goal of discovering useful information |
|---|---|
| Represents | data analysis (Q1988917) |
| Data type | Item |
| Domain | scholarly article (Q13442814)
scientific publication (Q591041) scholarly work (Q55915575) |
| Example 1 | Scientific article XXX→multivariate statistics (Q1952580) |
| Example 2 | Book XXX→qualitative data analysis (Q104773677) |
| Example 3 | randomized controlled trial (Q1436668)→survival analysis (Q543310) |
Motivation
editI did not find the right property to describe the method used in an academic work. This could be useful in the long run to make queries about scientific items and evaluate their quality for example.
(Maybe "uses data analysis method" is a better phrasing for the property name).
Jeanne Noiraud (talk) 21:25, 4 November 2024 (UTC)
Edit : I have listed existing properties on methods and my other property proposals here Wikidata:WikiProject_Wikidata_for_research/Data_models/Social_science_results#Properties_for_methods
Jeanne Noiraud (talk) 22:48, 5 November 2024 (UTC)
Discussion
edit
Comment I believe main subject (P921) could be used for this now (although the label maybe doesn't quite fit). Would it make sense to have a more general "method" property though? ArthurPSmith (talk) 20:58, 5 November 2024 (UTC)
- Hi, thank you for your comment.
- Some scientific work are discussing methodologies themselves, those papers can have data analysis method as their main subject (P921). But the other would be simply be using these methods, while their subject are behaviours, organisations, diseases...
- I think it would be better to have something more specific than "method". We usually make the distinction between the research design (experiment, case study...), the data collection method (questionnaire, observation, physical measures...) and the data analysis methods (qualitative analysis, statistic, comparative analysis...). I plan to make other proposals for these.
- Having details about this for each research item would be useful to fit open science best practices about metadata. Jeanne Noiraud (talk) 22:15, 5 November 2024 (UTC)
- There are already more than 31 million statements in wikidata using main subject (P921), so it would unlikely be able to support use cases involving SPARQL queries. On the other hand, having a property that is too specific or narrow in scope may end up being rarely used, as in heat treating (P6212). A compromise is to have a new property such as "analysis method", with study type (P8363) as Qualifier if needed. --Zhenqinli (talk) 23:13, 5 November 2024 (UTC)
- Hi, thanks, I missed study type (P8363), it would be a good fit for what I call "research design" in my jargon. Actually I would rather put "analysis method" as a qualifier for study type (P8363) rather than the other way around. But I still think it can be useful on its own as well. Jeanne Noiraud (talk) 16:06, 6 November 2024 (UTC)
- Sounds good. By the way, I see that you have another proposal Wikidata:Property proposal/Use data collection instrument. Just wonder if it could be served instead by the existing property uses (P2283)? In fact, you may browse the instances and subclasses of Wikidata property related to method, methodology or mode of actions (Q119301849) for other existing/available Wikidata properties related to methods. --Zhenqinli (talk) 15:27, 7 November 2024 (UTC)
- Hi, thanks, I missed study type (P8363), it would be a good fit for what I call "research design" in my jargon. Actually I would rather put "analysis method" as a qualifier for study type (P8363) rather than the other way around. But I still think it can be useful on its own as well. Jeanne Noiraud (talk) 16:06, 6 November 2024 (UTC)
- There are already more than 31 million statements in wikidata using main subject (P921), so it would unlikely be able to support use cases involving SPARQL queries. On the other hand, having a property that is too specific or narrow in scope may end up being rarely used, as in heat treating (P6212). A compromise is to have a new property such as "analysis method", with study type (P8363) as Qualifier if needed. --Zhenqinli (talk) 23:13, 5 November 2024 (UTC)
- Comment I think that uses (P2283) could be used instead of creating a new property. PAC2 (talk) 21:02, 10 November 2024 (UTC)
- @Jeanne Noiraud: please refer to actual items in your examples instead of "Scientific article XXX". When properties get created, we set the properties examples of a new property based on what's written in the proposal and that's not possible without concrete items. Concrete items make it also generally easier to reason about how well the property serves the specific role in which it's used instead of only thinking about it generally. ChristianKl ❪✉❫ 12:40, 13 November 2024 (UTC)
- While both main subject (P921) and uses (P2283) could theoretically contain this information, both of those are overly broad. I think it can be useful to have more specific properties in cases like this. ChristianKl ❪✉❫ 12:50, 13 November 2024 (UTC)
