carpentries / assessment Goto Github PK

From @maneesha on May 7, 2018 16:42

The table showing how many countries had their first workshop each year has an index of mixed types (ints for each year, plus string 2018Q1). How do we sort this by year?

Copied from original issue: carpentries/assessment#185

From @karenword on March 1, 2018 20:38

It would be great to know how or whether any of the changes that were made had an impact on outcomes!

Copied from original issue: carpentries/assessment#159

From @maneesha on June 8, 2018 15:58

For next report -- We currently have a table & visualization showing how how many workshops each badged instructor has taught. For those who have never taught, this should filter out those who have been badged recently (last 3 mos? 6 mos) and may not have had a chance to teach yet.

Copied from original issue: carpentries/assessment#204

In the analysis report, the figure showing the different options for "Involvement in Carpentries Community Post-Workshop" is missing the category of "Became a Carpentry Instructor", which is in the raw data.

Thank you for producing this report - it's so useful for us to show the ongoing benefit of Carpentries to funders and stakeholders. 👍

From @maneesha on May 17, 2018 20:47

For projections of number workshops for 2018, I took the proportion of 2017 workshops run in 2017 Q1 and applied it to the 2018 Q1 workshop count. This includes country specific and carpentry specific projections (so 2018 projections for USA workshops are based on 2017 USA data).

For now, the same thing is done for attendance data. The proportion of 2017 attendees who came in 2017 Q1 was applied to the number of attendees we saw in 2018 Q1.

The attendance projection for 2018 does not directly account for the number of workshops or class size. Should we do this? If so how do we do this?

Copied from original issue: carpentries/assessment#201

From @tracykteal on March 23, 2018 15:55

In the long term survey, for the perception of workshop impact, could you add the legend, with the bold words that match the ones in the figure, along with the question, like in the Draft Data Carpentry post-workshop survey? That one works great, and would be good here too.

Copied from original issue: carpentries/assessment#167

The pdf of the post-workshop survey (https://github.com/carpentries/assessment/blob/master/learner-assessment/surveys/swc_postsurvey_current.pdf) seems to be a copy of the pre-workshop survey, including the title "Software Carpentry Pre-Workshop Survey". I don't think this is intended

Can we add the link below to the area at the top of the repo that's meant for descriptions, links, etc?

https://carpentries.github.io/assessment/

Also, I wanted to assign an issue to @karenword and Kelly Barnes but was not able to. Do they have access to this repo?

From @lexnederbragt on June 15, 2018 8:33

Figure 10: Instructor teaching frequency is not very informative due to the extreme outliers. Maybe add a copy of the figure with outliers removed so the y-axis becomes a smaller range?

Copied from original issue: carpentries/assessment#206

From @maneesha on Apr 26 2018:

The workshop administrator (self organized or other administrator) has inaccurate/inconsistent data. @maneesha is working on fixing this with the Workshop Administrators.

copied from: carpentries/assessment#174

From @maneesha on April 26, 2018 18:5

In a Jupyter notebook we can have interactive graphs where users can do things like filter data, choose the graph layout, etc.

Is it possible to embed this in an html page outside of a Jupyter notebook?

Copied from original issue: carpentries/assessment#176

From @maneesha on June 8, 2018 15:31

We have a map that shows for each country that has ever hosted a Carpentries workshop, what year their first workshop was in.

For next report I would like to show sustainability - how many countries are "one and done" and how many have hosted at least one workshop a year since their first year. Ideas on how to visualize this are welcome.

Sample current graphics (data may change):

Year each country hosted first workshop:

Workshops by year for most active countries

Copied from original issue: carpentries/assessment#203

I think it would encourage learners to fill out the survey if they could see how long it is or a progress bar in the survey. I've noticed people at my workshops clicking out of the survey at the last minute, when they're close to being done, since they're unsure how much longer it will take.

Just a thought!

From @maneesha on May 17, 2018 19:6

See:

https://github.com/carpentries/assessment/blob/master/programmatic-assessment/workshops/outputs/sample_report.pdf

Table formatting gets thrown off.

Tables do render well in html:

http://htmlpreview.github.io/?https://github.com/carpentries/assessment/blob/master/programmatic-assessment/workshops/outputs/sample_report.html

Copied from original issue: carpentries/assessment#199

From @maneesha on May 7, 2018 17:17

We have charts and tables showing counts of inperson and online training events.
We also have a table of completion rates by country. Combine this to show the difference in completion rates by online vs. inperson.

Copied from original issue: carpentries/assessment#191

I'm curious as to if it would be possible to customize the "Software Previously Used" question to include software specific to the workshop being taught, if the workshop is a Data Carpentry one.

For example, as an instructor teaching the Geospatial Workshop, it's very important for me to know if my students have used ArcGIS/QGIS before, which is a point-and-click GIS software. This helps me structure my curriculum to account for my learners' needs.

I realize this might not be the goal of this assessment, so feel free to let me know it's not. Perhaps this can be solved by including an option in Q8 ("How often do you currently use any of the following?) for "Specialized disciplinary software with a point and click interface (ArcGIS, OpenRefine, etc)"?

From @maneesha on May 7, 2018 16:16

When images are exported (via download as html or other formats) they are given generic file names (e.g. output_19_2.png ). Is there a way to specify the filename in Jupyter?

Copied from original issue: carpentries/assessment#180

I'm working up a clone of the DC survey for our HASS workshop and I realised that the responses skew very heavily towards engineering/science specifics and leave off a generic "humanities" for us HASS folk.

Agricultural or Environmental Sciences
Ancient History
Anthropology
Biomedical or Health Sciences
Chemistry
Civil, Mechanical, Chemical, or Nuclear Engineering
Computer Science or Electrical Engineering
Criminology
Earth Sciences
Economics or Business
Education
English
Genetics, Genomics, or Bioinformatics
Geography and Planning
High Performance Computing
Humanities
Indigenous Studies
International Relations
International Studies
Law
Library and Information Science
Mathematics or Statistics
Media, Music, Communication and Cultural Studies
Medicine
Modern History
Organismal Biology (Ecology, Botany, Zoology, Microbiology, etc.)
Philosophy
Physical Sciences
Planetary Sciences (Geology, Climatology, Oceanography, etc.)
Politics
Psychology or Neuroscience
Security Studies
Social Sciences
Sociology
Space Sciences
Other (please specify)

Is the list unioned with our departments here. Might be worth thinking about for the next round of the survey.

In the README it says Assessment projects are led by Kari. I believe @karenword and I decided that the instructor training team would lead assessment moving forward, so I wanted to make sure that's true, and if so, ask that we update this README.

From @maneesha on May 7, 2018 17:14

Since people have 3 months to complete their certification towards a badge, no one who did training in 2018 Q1 would be expected to have completed their certification.

Copied from original issue: carpentries/assessment#190