Decide Next Steps

[timothyfraser]

What topics need more learning checks?

• [ ] Midterm Review Doc --> new versions of old questions, new questions (Statistical Process Control, Lifespan Distributions, Fault Trees) --> ready by end of week
• [ ] RSM: Learning Checks --> ready by November
• [ ] Probability Overhaul --> you read our probability section, honest take on what matters, what doesn't, what is needed. --> no firm deadline

[timothyfraser]

@dj297

• Multiple Choice Questions (Apply Six Sigma rules to Statistical Process Control diagrams)
  • 10 questions, 4 options per question.
  • Some can be calculation.
• Short Calculation Questions (Conditional Reliability given...., Failure Rate, probability at time t for the weibull, or exponential)
  • 5 questions
• Terrifying Questions (Fault Trees, Block Diagrams, Chisquared Calculation, etc.)

I'm going to send you...

• a shared Google doc link to put your questions, code, anything in
• link to slides
• links to textbook chapters
• Chapters Covered in this Exam
  • Module 4: System Reliability
  • Module 5: Statistical Process Control
  • Module 6: Other Useful Charts
  • Module 7: Useful Life Distributions
  • Module 8: Useful Life Distributions II
  • Module 9: Fault Trees
• links to past review doc and solutions

Partial Deadline: Wednesday Full Deadline: Friday

[timothyfraser]

@dj297 Hi Devanshi, all material is now uploaded. Let me know here if you have any questions! Thanks for your help!!

[dj297]

@timothyfraser Hi Tim,

I've added a few questions and am in the process of adding the remaining. Feel free to share any suggestions if you happen to review the document.

[timothyfraser]

Hi Devanshi @dj297 , Thanks for these questions. These are really nice! If you can add any remaining questions by noon EST on Friday, that will be very helpful! Some questions regarding bootstrapping, confidence intervals, etc. would be nice. Best, Tim

[dj297]

Hi Tim @timothyfraser,

I've added the remaining questions to the document. Feel free to add any edits/ suggestions if you happen to review the document.

[timothyfraser]

Hi Devanshi @dj297 , These are fabulous questions, and a great addition. Thanks for your hard work. I'll go ahead and combine these with my other questions! Much appreciated! Let's touch base next week about next steps.

[timothyfraser]

Hi Devanshi,

Here's my request!

Course Design Priorities

• capable R coders
• do six sigma analyses in R
• do system reliability analyses in R
• can write functions.
• do tidyverse functions (eg. dplyr, ggplot2)
• fabulous at exponential and weibull distributions

Not priorities:

• soul-destroying data wrangling
• don't need them to use tidyfault
• don't need them to use obscure probability rules
• don't need them to do ridiculously long calculations if it could be wrapped in a function.

Request

Could you write up a list of these 2 types of things:

• https://timothyfraser.com/sigma/

1. Sections that can be cut (eg. density() )

• [ ] section 1
• [ ] section 2
• [ ] stuff

2. Skills or functions that need some help.

• [ ] section 1
• [ ] section 2
• [ ] stuff

3. Learning Check Ideas if you have them.

• [ ] section 1
• [ ] section 2
• [ ] stuff

[dj297]

@timothyfraser Hi Tim,

I haven't had the chance to dedicate time to reviewing the sections because I'm departing from India this week. I've been occupied with some last minute formalities and spending time with family and friends. I'll make an effort to make progress while I'm on the go and will provide you with an update shortly.

[timothyfraser]

Hi @dj297 Devanshi,

That's totally fine. Wishing you safe travels with your move.

[dj297]

@timothyfraser I'm adding a few suggestions:

1. Skill Visualisation with ggplot in R:
   • [ ] Removing learning check from gg-lot visualization (7, 8)
   • [ ] Also learning check 10 - facet wrap (haven’t used it personally so I’m not sure how crucial it is).
2. Skill: Failure Modes and Effects Analysis in R:
   • [ ] If we have data for severity, occurrence, and detection, we could show tabulations for just one of these and provide a function for calculating RPN. This would reduce the number of sections, and learners could create tables for each parameter as learning checks for practice.
3. Workshop 2 - Distributions and Descriptive Statistics:
   • [ ] Learning Check 1 may not be necessary for Distributions (Workshop 2)
   • [ ] Descriptive Statistics - You could consider mentioning how to calculate one or two statistics, such as mean and median. Since we've already listed all the functions, learners can practice working out other statistics through learning checks using the provided info.

Apologies for the delay, I'm currently working on more sections and will keep adding to the list. Your thoughts on whether these suggestions make sense would be appreciated. If you think discussing this further in a meeting would be helpful, just let me know when you're available.

[timothyfraser]

Hi Devanshi, @dj297 Thanks for this. These are helpful. The point about facet_wrap is funny - I never thought of that! Can you finish your list, and then we'll do one final meeting to discuss?

Best, Tim

[dj297]

@timothyfraser Adding a few more suggestions, and will complete the list hopefully by next week. I hope these make sense and we can discuss further during the final meeting.

Workshop 4: System Reliability

• [ ] Use the New York Car example for initial understanding and introduce the propeller example for reinforcement.
• [ ] Can provide functions for one or two types of failure rates to give a head-start. Additional functions can be mentioned in the appendix if needed.
• [ ] Offer one example of units for better comprehension, as they've already learned how to make functions.
• [ ] Convert units through learning checks, as it might be challenging to remember making conversion functions.
• [ ] Include an R code/function for calculating series/parallel system reliability in section 4.5 to enhance understanding.
• [ ] Turn the cumulative probability of replacement function (fr_f()) into a learning check or workshop assignment in section 4.6.

Workshop 5:  Statistical Process Control in R

• [ ] Unsure of the significance of content post-5.3.2; Mentioning factors may be moved to the appendix or introduced through learning checks for clarity.

Workshop 6: Indices and Confidence Intervals for Statistical Process Control in R

• [ ] 6.2 These are very useful sections; can consider converting the later parts into a group activity or a structured learning check.
• [ ] For 6.4, we could demonstrate calculating the CI for one index and then encourage participants to tackle the rest independently.

\Workshop 7 : Useful Life Distributions (Exponential)

• [ ] Unsure whether we can jump into building the Chi-squared function in 7.5.4 without bringing up the earlier section (7.5.3) or if it's better to put it in the appendix (7.5.3). It's essential to get why we're making it a function, but going through all pf it might overshadow the focus on the actual function creation.
• [ ] The sections following 7.5.4 are important for review but should be also prioritized for an in-depth discussion of these sections in class (the coding part fall all the estimations) - helps understand why making a function helps in the long run. Workshop 7 is long and can be overwhelming and so it can be presented in parts.

[timothyfraser]

@dj297 Hi Devanshi, These are great ideas. Thank you for your hard work. I'd like to propose that we meet one last time after Thanksgiving (my schedule next week is a bit packed). Would love to ask a few clarifying questions, get your thoughts. Can of course be flexible with time since you are starting your new job.

[dj297]

@timothyfraser Hi Tim, Happy Thanksgiving! I was also occupied with finding housing and paperwork but will finish reviewing sections soon. Will update you shortly and schedule a meeting. Hope you're having a great time!