assignments.Rmd

---
title: "Bayesian Data Analysis course - Assignments"
date: "Page updated: `r format.Date(file.mtime('assignments.Rmd'),'%Y-%m-%d')`"
---

You are free to use these assignments in self study and other courses
(CC-BY-NC 4.0), but please do not publish complete answers online.

If you are a student on this course, you are allowed to discuss assignments with your friends, but it is not allowed to copy solutions directly from other students or from internet. You can copy, e.g., plotting code from the course demos, but really try to solve the actual assignment problems with your own code and explanations. Do not share your answers publicly. Do not copy answers from the internet or from previous years. We compare the answers to the answers from previous years and to the answers from other students this year. All suspected plagiarism will be reported and investigated. See more about the [Aalto University Code of Academic Integrity and Handling Violations Thereof](https://into.aalto.fi/display/ensaannot/Aalto+University+Code+of+Academic+Integrity+and+Handling+Violations+Thereof).

## Weekly assignments

There are 9 weekly assignments (two of them have two weeks to
submit). Assignments are linked from [the schedule](Aalto2021.html#Schedule_2021) and can also be found from [git repo assignments
folder](https://github.com/avehtari/BDA_course_Aalto/tree/master/assignments). The deadline days for the assignments are given
in [the course schedule](Aalto2021.html#Schedule_2021).

- The assignments are introduced on Mondays.
- There are chat streams **#assignment1** etc. you can ask questions about the assignments. Other students and TAs can answer these. There is no guaranteed response time. These streams are best for questions that are likely to have relatively simple answer.
- There are [TA sessions](#TA_sessions) for getting help. These sessions are not obligatory. These sessions are useful if you think you need help that requires a bit more discussion. The questions are answered during the TA session time (if there are too many questions, they may be answered in the chat or next TA session).
- You can use [R markdown template](https://github.com/avehtari/BDA_course_Aalto/tree/master/templates/) or [Quarto template](https://github.com/avehtari/BDA_course_Aalto/tree/master/templates/) for the report. **You can also use any other software**, but please follow the outline and ideas in the template [also available as PDF](https://github.com/avehtari/BDA_course_Aalto/tree/master/templates/)
- RStudio has a visual R markdown editor which makes it very easy to edit R markdown and Quarto documents as it looks more like the end result and has handy toolbar and menus for formatting (e.g. section headings, bolding, lists, figures)
- Students return their answers to
[peergrade](https://www.peergrade.io) by the end of the week (hand-in
period). The deadlines are on Sundays 23:59. We can't accept late submissions due to the peergrading process.
- We highly recommend to submit all assignments Friday before 3pm so that you can get TA help before submission. As the course has students who work weekdays (e.g. FiTech students), the late submission until Sunday night is allowed, but we can't provide support during the weekends.
- After this, each student reviews 3 random other students'
answers and provides feedback via online rubric form in peergrade during Monday to
Wednesday 23:59 (peer grading period).
- After peergrading, each student reflects on the feedback (reactions, e.g. not helpful/helpful) before Sunday 23:59.
- If a student receives inappropriate feedback/grading or reaction, they may "flag" it
for TAs to check from Wednesday to Sunday (flagging period).
Strongly conflicting feedback/gradings are also manually checked by TAs (after
flagging period). As a guideline, the TA responsible for that particular
assignment will resolve these flags within one week from the date when the feedback was closed. Sometimes, especially if the flag resolving is escalated it may take more time. Flags that are added after one week from the when the feedback was closed will be resolved in the end of the course.
- Peergrade alerts: If you are worried that you forget the deadlines, you can set peergrade to send you email when assignment opens for submission, 24 hours before assignment close for submission, assignment is open for reviewing, 24 hours before an assignment closes for reviewing if you haven't started yet, someone likes my feedback (once a day). Click your name -> User Settings to choose which alerts you want.

NOTE: The assignment instructions can be updated during the course, and
individual assignments are not guaranteed to be up to date until Monday
8am of the hand-in period of the corresponding assignment week (See
the deadlines below, e.g. the first assignment is not guaranteed to be up to
date until 13.9 at 8am).

Report all results in a single, **anonymous** *.pdf -file. Do not include your name, student id, or other information that can identify you (Peergrade knows who submitted what). You can choose the file name of freely. Submit the report
it in [peergrade](https://www.peergrade.io). Submitting empty or
almost empty pdf files is not allowed (this will be handled as a
violation of academic integrity as it is harming other
students). Include also any source code to the report (either in
appendix or embedded in the answer). By anonymity it is meant that the
report should not contain your name or student number. In addition to
the correctness of the answers, the overall quality and clearness of
the report is also evaluated.

The assignments are mostly solved using computer (R or Python). Related
demos for each assignment are available in the course web pages (links
in Materials section). See [TA sessions](#TA_sessions) for getting help.

Overall recommendation for many students is to include more text
explanations as writing that text is active way of learning, and also
it's easier to see if you have understood the concepts correctly. If
you make minimal reports, you will also get minimal feedback and you
learn less. The main purpose of the course is to learn, and learning
is more effective if you do it in active and interactive way.

For those who are self studying or who miss several assignment
hand-ins, there is a possibility to hand in all assignments once in
January (no need to ask permission for this as this is automatically
granted for everyone). We still strongly recommend students taking the
course to follow the weekly submission deadlines to more evenly spread
the work for them and the reviewers during the course.

## Peergrade

Signin into peergrade with link provided at [MyCourses](https://mycourses.aalto.fi/course/view.php?id=32457)

## Peergrading / feedback

Peergrading is called giving feedback in peergrade.io. Giving feedback
is good term as it is not meant to be just grading wrong vs. correct
answers. Please provide also feedback on the presentation. Part of the
course is practicing making data analysis reports. By providing
feedback on the report presentation other students can learn what they
can improve or what they already did well. You should be able to
provide constructive or positive feedback for all non-empty reports,
even if there is nothing to say about the technical correctness of the
answers. If you all provide more feedback, you will all learn more
about good data analysis reports. Be kind.

## Assignment scoring

Points are given from both *submitting an assignment* and *giving
feedback*. *Submission performance* gives 70\% and *feedback performance*
30\% of the total score. Only those students who returned a given
assignment are allowed to give feedback to the answers submitted by others.

You don't need to get 100% score from the submission or feedback to
get grade 5. Thus, one should not worry about getting a 100\% score. A
critical reviewer will also affect other students' grading. In
addition, a question can always be flagged for course personnel to
check.

We have noticed in previous years that some students return an empty
pdf to peergrade to still get points from giving feedback. This is not
allowed. If a student returns an empty pdf or it is obvious from the
pdf that the student hasn't done any effort for the assignment, the
student is not allowed to give feedback. The course staff will monitor
returned assignments, and negative points will be awarded to empty assignments.
If any student is reviewing an empty pdf, please contact course staff.

### Submission score

After peer grading is over for an assignment, each submission receives a
*submission score*. This score is between 0\% and 100\% depending on how
well the peers evaluated the submission. This is mostly about the
technical correctness and small part about the presentation.

The numerical score for each criteria is normalised to be between 0\%
and 100\%, where 0\% is given to the worst, and 100\% to the best
possible answer. As an example, in a Scale-question with 5
levels, the different answers will count 0\%, 25\%, 50\%, 75\%, 100\%.

The total submission score is calculated for a submission by averaging
the scores from each of the peers. The score from an individual peer
evaluating a submission is obtained by averaging over the criteria in
the feedback rubric.

Within each assignment, there are two sections *Basic requirements* and
*Overall quality of the report*. These segments are weighted so that
they each provide 7.5\% of that assignment's total score. The remaining
85\% is distributed evenly to the rest of the sections. Within each
section, the points are distributed evenly among each question.

If a student has flagged some of their received evaluations and a TA has
accepted the flag and overruled the original evaluation, then TA's new
evaluation is used for computing the score instead. If a teacher
overrules a specific evaluation by one student, that answer is taken to be
the correct one for that part of that submission. This means that
the evaluations by other students on that part are disregarded in the final
score of the submission.

There are 9 assignment rounds in total. Each round has different
weighting when computing the total score. The assignment instructions
list max points for each assignment as 3, 3, 9, 6, 6, 6, 6, 6,
3. These translate (approximately) to weights 6%, 6%, 19%, 12%, 12%,
12%, 12%, 12%, 6%.

### Feedback score

The *feedback score* depends on the reactions, such that a student's
feedback score on an assignment is an average of the reactions
received. If no reactions have been given to the student's reviews, no
feedback score is computed and it will simply show as "-".

When you provide feedback, you can provide constructive and positive
feedback n how to improve presentation or tell what is already very
good. By providing that kind of feedback you are more likely to
receive higher feedback score.

The following are reasons to react positively to feedback:

- Reviewer giving you tips on how to improve the presentation of the report or specific answer.
- Reviewer making the effort of finding and pointing out where your bug originates from.

Be kind when reacting to to the feedback. If you didn't like the
feedback, think first a moment, and then re-consider whether it was
just the wording that got you angry and maybe there is something
useful in the feedback.

Note that the baseline feedback reaction is "Somewhat useful. Could be
more elaborate." which corresponds to 0.5 points or as Peergrade shows
it to 50%.  It might make you feel better if you consider that the
middle option is 100%, as then you have already provided useful
feedback which is the goal, and then the higher categories "Very
useful" and "Extremely useful" would be 150% and 200% and you can
consider receiving bonus points.

### Bonus points

In addition to the peergrade score, one can get bonus points from
course chat activity and by answering a time usage questionnaire on the course page in [MyCourses](https://mycourses.aalto.fi/course/view.php?id=32457).
Course chat activity and the questionnaire are not included in the maximum assignment
score and thus are not required to receive a full 100\% score.

Other students' course chat activity will not affect one's resulting
grade, i.e. there is no need to try to perform in the course chat or to take any part in it at all.

After each assignment, a voluntary anonymous questionnaire is opened on the course page in] [MyCourses](https://mycourses.aalto.fi/course/view.php?id=32457) for the peergrading feedback period (Monday 8:00 to Wednesday 23:59).

## Be polite

Remember to be polite when peergrading, giving feedback, and reacting to
feedback. Do not spend a lot of time fighting over the grading of one question
from one student. If you don't agree you can submit a flag or in extreme case
contact the TAs. Also, keep in mind that in most of the cases, which
we've seen, the students have been fighting for points which have less
than a 1/1000 effect on the final score. Long fight over that is not
worth the time and energy.  If you get feedback which makes you angry, breath and wait a
moment before unleashing your anger back. We ask you to honor the
system and be polite to your peers.

## TA sessions

You can get help for the assignments by asking in the course chat from other
students or in weekly TA sessions by asking TAs. The sessions are
voluntary.

- There are chat streams **#assignment1** etc. you can ask questions about the assignments. Other students and TAs can answer these. There is no guaranteed response time. These streams are best for questions that are likely to have relatively simple answer and thus are likely to be answered before the next TA session.
- There are [TA sessions](#TA_sessions) for getting one-on-one help. These sessions are not obligatory. These sessions are useful if you think you need help that requires a bit more discussion. The questions are answered during the TA session time (if there are two many questions, they may be answered in the chat or next TA session).

There are three TA sessions each week; please see [the course
schedule](Aalto2021.html#Schedule_2021) for more details on the meeting
times. In the TA sessions, you can get one-to-one help with your
assignments and project work.

During the TA session you can get help in the following forms:

- Written communication on *the course chat*: you will chat with a TA using the “direct messages” feature on *the course chat*. You can also, for example, share code snippets and equations through chat direct messages if it helps.
- Oral communication on Zoom: you will chat with a TA using a video conference on Zoom. You can also use, for example, screen sharing on Zoom if it helps.

We will use the stream **#queue** in the course chat to coordinate everything. We announce there when the TA session starts. Then you can write your help request there, describing in sufficient detail exactly what is the problem with which you would need help (see below).

Once a TA is free and your question is the first request in the queue, a TA will mark it with a check mark reaction. Then the TA will contact you and help with your problem. Finally, once the problem is solved, the TA who helped you will delete your request from the queue.

### Getting help via the course chat

- Login with Aalto account to Zulip course chat (link in [MyCourses](https://mycourses.aalto.fi/course/view.php?id=32457))
- Go to the chat stream #queue.
- Write a help request (see below), starting with the keyword “Chat”.
- A TA will send you a direct message on Chat.
- You will discuss through direct messages until your problem is solved, and then the TA will close the discussion and delete your help request.

Please do not send direct messages to TAs without going through the above protocol. If you have follow-up questions later, please put a new request to the queue.

### Getting help via Zoom

- Follow [these](https://www.aalto.fi/en/services/zoom-quick-guide) instructions to install Zoom and to sign in to Zoom.
- Open Zoom, and make sure your video and audio are configured correctly. Create a new Zoom conference call, and copy the meeting URL.
- Go to the Chat stream #queue.
- Write a help request (see below), starting with the keyword “Zoom”, and end it with the Zoom meeting URL.
- A TA will click on the meeting URL to join the Zoom conference that you created.
- You will get help until your problem is solved, and then the TA will close the call and delete your help request.

### What to write in the help request?

Your help request should contain a concise summary of exactly what kind of help you would need. Ideally, after reading the help request, a TA should be able to already have an answer for you.

Try to describe what is the problem, what you have tried, what you already know, and exactly what is the relevant part of the code. Please highlight the important parts. Here are some fictional examples of good help requests:

---

**Chat**: I am not able to get the correct solution to assignment 2 part c) by using the test data. According to the assignment pdf, the correct solution is 0.4511238, but I am getting the answer 0.3771946. The code for my function is:

`EXAMPLE CODE 1`

I thought that I have an error in ..., so I tried changing ... to ..., but then I got a different wrong answer, 0.4285443. I think the error may be on line 3 in the code, but I do not know how to fix that.

---

**Zoom**: I am trying to install the R package ... on my personal laptop and I am getting the following error:

`EXAMPLE ERROR`

My operating system is ..., I have version ... of R installed and I am using RStudio. I tried googling the error but was not able to solve the issue. Zoom meeting link: https://aalto.zoom.us/j/XXX

---

### Acknowledgements {.toc-ignore}

TA session instructions above have been copied from [Programming
Parallel Computers by Jukka Suomela](http://ppc.cs.aalto.fi/2020/lab/)
with CC-BY-4.0 license.