Assignment 7

S1

Basic requirements

Q1

Can you open the pdf and it’s not blank?

  • No
  • Yes

Q2

Is the report anonymous?

  • No
  • Yes

S2

Linear model: drowning data with Stan

Q3

Is the source code included?

  • No
  • Yes

Q4

Is the full resulting modified Stan model code presented in the report?

  • No
  • Yes

Q5

Fix #1: Is there a fix for … ?

  • It has not been discussed that …
  • It has been discussed that …, but there is no fix presented for it or the fix is clearly wrong
  • There is a fix presented for …, that clearly solves the problem

Q6

Fix #2: Is there a fix for …?

  • It has not been discussed that …
  • It has been discussed that …, but there is no fix presented for it or the fix is clearly wrong
  • There is a fix presented for … that clearly solves the problem

Q7

Fix #3: Is there a fix for …?

  • It has not been discussed, that …
  • It has been discussed that …, but there is no fix presented for it or the fix is clearly wrong
  • There is a fix presented for … that clearly solves the problem

Q8

Is there a suitable numerical value of approximately … presented for \(\sigma_{\beta}\) (or … for \(\sigma_{\beta}^2\))?

  • No
  • Yes

Q9

Does the report discuss and correctly present how the desired prior can be implemented in the model code?

Example implementation:

  • No
  • Yes

Q10

Does the report discuss and also correctly present a prior for the intercept?

Example implementation:

  • No
  • Yes

S3

Hierarchical model: factory data with Stan

Q11

Separate model: Is the model described using mathematical notation and the difference to other models described in words?

  • No equations and no description
  • Description but no equations
  • Equations but no description
  • Equations and description

Q12

Separate model: Is there a related Stan implementation (N.B. same implementation may be used for multiple models)?

  • No Stan model implemented
  • Stan model implemented, but it seems clearly wrong or broken
  • Seemingly valid Stan model implemented

Q13

The following histograms are used as a reference in the grading of the separate model:

Separate model: Discussion about question c part i, the posterior distribution of the mean of the quality measurements of the sixth machine

hidden responses

Q14

Separate model: Discussion about question c part ii, the predictive distribution of a quality measurement from the sixth machine

hidden responses

Q15

Separate model: Discussion about quection c part iii, the posterior distribution of the mean of the quality measurements of the seventh machine

hidden responses

Q16

Separate model: When using the prior of Normal(0,10) and Gamma(1,1) for mu parameter(s) and sigma parameter(s), the posterior for mu for machine 1 has 90% credibility intervals close to: …

  • No or incorrect answer
  • Answer is only partially correct
  • Answers look correct

Q17

Is the pooled model described using mathematical notation and difference to other models described in words?

  • No equations and no description
  • Description but no equations
  • Equations but no description
  • Equations and description

Q18

Pooled model: Is there a related Stan implementation (N.B. same implementation may be used for multiple models)?

  • No Stan model implemented
  • Stan model implemented, but it seems clearly wrong or broken
  • Seemingly valid Stan model implemented

Q19

The following histograms are used as a reference in the grading of the pooled model:

Pooled model: Discussion about question c part i, the posterior distribution of the mean of the quality measurements of the sixth machine

hidden responses

Q20

Pooled model: Discussion about question c part ii, the predictive distribution of a quality measurement from the sixth machine

hidden responses

Q21

Pooled model: Discussion about question c part iii, the posterior distribution of the mean of the quality measurements of the seventh machine

hidden responses

Q22

Pooled model: When using the prior of Normal(0,10) and Gamma(1,1) for mu parameter(s) and sigma parameter(s), the posterior for mu for machine 1 has 90% credibility intervals close to: …

  • No answer
  • Answer is only partially correct
  • Answer looks correct

Q23

Is the hierarchical model described using mathematical notation and difference to other models described in words?

  • No equations and no description
  • Description but no equations
  • Equations but no description
  • Equations and description

Q24

Hierarchical model: Is there a related Stan implementation (N.B. same implementation may be used for multiple models).

  • No Stan model implemented
  • Stan model implemented, but it seems clearly wrong or broken
  • Seemingly valid Stan model implemented

Q25

The following histograms are used as a reference in the grading of the hierarchical model:

Hierarchical model: Discussion about question c part i, the posterior distribution of the mean of the quality measurements of the sixth machine

hidden responses

Q26

Hierarchical model: Discussion about question c part ii, the predictive distribution of a quality measurement from the sixth machine

hidden responses

Q27

Hierarchical model: Discussion about question c part iii, the posterior distribution of the mean of the quality measurements of the seventh machine

hidden responses

Q28

Hierarchical model: When using the prior of Normal(0,10) and Gamma(1,1) for mu parameter(s) and sigma parameter(s), the posterior for mu for machine 1 has 90% credibility intervals close to: …

  • No answer
  • Answer is only partially correct
  • Answers look correct

S4

Overall quality of the report

Q29

Does the report follow the formatting instructions?

  • Not at all
  • Little
  • Mostly
  • Yes

Q30

In case the report doesn’t fully follow the formatting instructions, specify the formatting instruction that is not followed. If applicable, specify the page of the report, where this difference in formatting is visible.

Q31

Please provide also feedback on the presentation (e.g. text, layout, flow of the responses, figures, figure captions). Part of the course is practicing making data analysis reports. By providing feedback on the report presentation and 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 nonempty reports, even if there is nothing to say about the technical correctness of the answers. You can also provide feedback on code.