search

UBC-MDS / speed_dating_analysis

This repo is for a group project for analyzing speed dating data set for MDS-522
Other
0 stars 0 forks source link

Which question to ask, what model to use for it, and what assumptions we are making for that model #2

Open wenyunie opened 1 year ago

wenyunie commented 1 year ago

We chose the speed dating data set, we found the data set interesting but also very challenging, because of the potential cofounding and how it is difficult to solve these problems with the currently learned models.

  1. Each rating is an observation and see what's the predictive power of each attribute on if the overall rating is Yes or No
  2. Each rate person is an observation and what's the predictive power of each attribute on if the overall rating (we will find a way to calculate the summed score from different raters on this person) 3.Self-rating v.s. other rating, level of analysis not decided
monazhu commented 1 year ago

Action items:

mishelly-h commented 1 year ago

Action items:

monazhu commented 1 year ago

@mishelly-h would you mind adding the data file and word doc for the data key, and add the .gitignore file for .csv and .doc/docx as well?

wenyunie commented 1 year ago

We chose the speed dating data set, we found the data set interesting but also very challenging, because of the potential cofounding and how it is difficult to solve these problems with the currently learned models.

  1. Each rating is an observation and see what's the predictive power of each attribute on if the overall rating is Yes or No
  2. Each rate person is an observation and what's the predictive power of each attribute on if the overall rating (we will find a way to calculate the summed score from different raters on this person) 3.Self-rating v.s. other rating, level of analysis not decided

Tiffany's reply on the main goal of this group assignment: https://ubc-mds.slack.com/archives/C24HUNRJL/p1699553458680469?thread_ts=1699492423.893409&cid=C24HUNRJL

wenyunie commented 1 year ago

2023.11.16 Lunch Break

Independent Variable: self rating v.s. other rating

Dependent Variable: an attribute of interest we are going to pick from the survey

question: Are self-rating and other-rating different for Attribute XXX in the context of speed dating?

rorywhite200 commented 1 year ago

We agreed to constrain the scope of our question to keep it simple. Our null hypothesis is that self-perception of attractiveness is accurate (there is no significant difference between self vs external attractiveness ratings). Our alternative hypothesis is that self-perception of attractiveness is inaccurate (there is a difference between self vs external attractiveness ratings).

monazhu commented 1 year ago

Task list:

monazhu commented 1 year ago

I realized we didn't discuss creating the renv file, as well as the references and README files.

@wenyunie, would you be okay with populating a renv file once you finish the preliminary analyses?

I can start populating a references file (I'll include what I needed to cite for the methods section) and work on what I can in the README project file for you guys to review.

wenyunie commented 1 year ago

I realized we didn't discuss creating the renv file, as well as the references and README files.

@wenyunie, would you be okay with populating a renv file once you finish the preliminary analyses?

I can start populating a references file (I'll include what I needed to cite for the methods section) and work on what I can in the README project file for you guys to review.

Yes, for sure.

mishelly-h commented 1 year ago

How do you guys want to do the citations? We could either setup a bibliography file or we can do it manually since this project is not that big and we do not have that many references. is that the references file that you mentioned?

monazhu commented 1 year ago

Oh good call - I was just gonna manually format it into APA (I think the methods section only has 2-3 references). Alternatively we can also share references on something like Zotero before porting it over. Rory and Michelle, since you two are gonna have more references to deal with, what would work better for you?

On Fri, Nov 17, 2023, 1:26 p.m. mishelly-h @.***> wrote:

How do you guys want to do the citations? We could either setup a bibliography file or we can do it manually since this project is not that big and we do not have that many references.

— Reply to this email directly, view it on GitHub https://github.com/wenyunie/speed_dating_analysis/issues/2#issuecomment-1817127968, or unsubscribe https://github.com/notifications/unsubscribe-auth/AB2DL32HFTTTPSPTMIYH3ATYE7JAZAVCNFSM6AAAAAA7DZLY4KVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTQMJXGEZDOOJWHA . You are receiving this because you commented.Message ID: @.***>

mishelly-h commented 1 year ago

I am perfectly fine to just do it manually since the project is not that big and figure out the automated way of referencing later. But if the anyone has strong preferences towards doing it automatically, I am open to do so.

rorywhite200 commented 1 year ago

Hello Michelle and all, I am happy to do this manually. APA style works for me.

monazhu commented 1 year ago

Okay sounds good! Give me 15 mins and I'll push what I have to the repo.

On Fri, Nov 17, 2023, 3:28 p.m. rorywhite200 @.***> wrote:

Hello Michelle and all, I am happy to do this manually. APA style works for me.

— Reply to this email directly, view it on GitHub https://github.com/wenyunie/speed_dating_analysis/issues/2#issuecomment-1817246989, or unsubscribe https://github.com/notifications/unsubscribe-auth/AB2DL34VYHTATCAC7YAWJDDYE7XLNAVCNFSM6AAAAAA7DZLY4KVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTQMJXGI2DMOJYHE . You are receiving this because you commented.Message ID: @.***>

monazhu commented 1 year ago

Hi everyone, just wanted to let you know that the following are up to date on my end:

I've also updated the README file to include a skeleton for the project file (for details, see pull request). I'm going to open a new branch for sharing the final report file (I'm planning to call it analysis_report.Rmd for now but feel free to change the name later). This is mostly to ensure that the parent/child document is working as expected.

wenyunie commented 1 year ago

Hi everyone, I finished the analysis and plot part. Can anyone make a review to the pull request and see it is approvable?