Hello, dear friend, you can consult us at any time if you have any questions, add WeChat: daixieit

Department of Statistical Sciences

STA305

Project Report

April 6, 2023

1. DOs

What to submit? When due?   Where?

1

Project Report as a single PDF le (1)

6 April 2023

Report, Crowdmark

2

R/Rmd codes

6 April 2023

R/Rmd, Crowdmark

3

Outputs, data sets etc. as a PDF le

6 April 2023

Output, Crowdmark

4

Video Presentation le

6 April 2023

Quercus

2. DON’Ts

2.1 E-mail submissions will NOT be accepted.

2.2 SUBMIT A SINGLE PDF DOCUMENT FOR THE REPORT. DO NOT SUBMIT descriptive texts in one document and graphs/tables etc. in another.

3. DETAILS in the following pages.

Project: General

1   An Overview

Each student  (no groups allowed) will plan and perform a homemade factorial ex- periment, or carry out an analysis of data from an observational study available on websites, using the guidelines given in the following. This includes:

a) designing an experiment, collecting or generating the data, and analyzing the data in R (for factorial experiment),

OR

b) description and data searched from various sources [other than the data used in the course] and analysis in R (for observational study).

You are encouraged to choose a topic of your interest for this project.

The report should not be longer than 4 pages including tables and plots, and should be submitted in one document.

An example of report layout is given at the end of this document. Feel free to use this report as a template.

Each student is required to submit the following four items ( Where do you submit what?):

Crowdmark

1. Your nal report as a single PDF le (due: 6 April 2023, 23:59 hr on Crowdmark). See the question box: Report.

2. R/R Markdown codes le (due: 6 April 2023, 23:59 hr on Crowdmark). See the question box: R/Rmd.

3. Supplementary materials(s) such as output le, along with necessary data files, etc. save them as PDF le(s) (due:  6 April 2023, 23:59 hr on Crowdmark).  See the question box: Output.

Quercus

4.   Video presentation of the objectives, methods, results,  and conclusions of your experiment (due: 6 April 2023, 23:59 hr on Quercus). See more on the format and size etc. in PART III.

2   Details

2.1    PART I. Guidelines for the project on factorial experi- ment and Observational Study

1a.  Number of factors under investigation should be more than two.  For example, you may develop a 24  factorial in blocks of size 8 or less, confound the interactions, preferably not the same set of interactions in all the replicates. Also use the common understanding “higher order interactions are either negligible or of lesser importance compared to the lower order interactions” . Main effects must not be confounded with blocks. You may also choose a fractional factorial design.

Other experimental designs could be a two factor factorials in Latin Squares, or as a split plot or strip plot design in RCBD or CRD, or even split-split plot design or other variants.

OR

1b.  Alternatively, you can present an Observational Study (OS), using the data on websites, but not the study presented in the course.

2.  Write a title for your experimental investigation or the observational study in the Report.

3. Write [Author: Your name, ID, specialization] in the Report.

2.2    PART II. Project Report

(Submit by 6 April 2023, 23:59 hr)

This document should have:

Introduction/Description (1/2 page maximum)

Introduce the need for this study, i.e., the rational behind this study, or, why and what did you plan to study/investigate?  Write the objectives of the study.  Cite any references used.

Materials and Methods (2 pages maximum)

Write in short the Materials and Methods you have pursued. This may include:

Experimental Design and Data

Describe the experimental design used — details of the factorial designs you generated or acquired, features of the designs, comment on how this design align with the ob- jectives you have set.  Cite any references used. Details of the data. How did you get the data randomly generated? found from a secondary source on the Internet? etc. Cite clearly the references. If you have arbitrarily inputted the data, write that as well.

Observational Study and Data

Most likely you will be using study described on a website. Describe the source of the study giving its url details of the study, time and location frame, list of variables: name and type of variables, treatment factor and a response variable.  Cite any refer- ences used.

Statistical Analysis

Details of the statistical methods used in the  analysis.   Cite  any references used. R/RStudio recommended for computations save the codes as *.Rmd. You may nd it more convenient to run the *.Rmd le on RStudio.cloud or JupyterHub, but they are not mandatory.

Results and Discussion (1 page maximum)

Describe and display your project ndings in text, tables and gures along the lines of the objectives (including whether the results support or contradict what you expected). Discuss an overall canvas of the study. Cite any references used.

Conclusion (1/2 page maximum including References)

Present in short the highlights of the study.

References (within 1/2 page of the Conclusion)

List the references in alphabetical order .

Format and size of the document:

Due date for all the submissions: 6 April 2023, 23:59 hr

Recommended font size is 12pt, but not less than 10pt.

Crowdmark

1.  Report in only PDF, no more than 4 pages, submit this le on Crowdmark.  This file should not contain R codes. Upload in the question box Report.

2. R/R Markdown codes le on Crowdmark. Upload in the question box R/Rmd.

3. Output le, along with necessary data files, etc. save them as PDF file(s) on Crowd- mark. Upload in the question box Output.

Quercus

4.   Video presentation of the objectives, methods, results,  and conclusions of your experiment (due: 6 April 2023, 23:59 hr on Quercus). See more on the format and size etc. in Section 2.3.

2.3    PART III. Video presentation.

(Submit by 6 April 2023, 23:59 hr, and on Quercus)

Your video must meet the following criteria:

a In the beginning of the video, you must clearly present your student ID along with

yourself.  The grading TA must be able to identify you and your student ID number. Failure to present your student ID will result in a 0 grade for the video presentation. We recommend that you update your Quercus profile with a picture where your face is clearly identifable.

b The presentation should not exceed 5 minutes. Any video beyond 5 minutes will not be

viewed by the grading TA, and will not be considered when marking. In the video you should describe the objectives, methods, results, and conclusions of your experiment.

c The video may be of any form, so be creative! For example, you may include clips of

yourself conducting the experiment while describing the experiment - beware that the clips will also count towards your 5 minute limit.

d Notes on video submission. If you are using a Mac, the Quercus media recorder submis- sion page may not work on your Safari browser. The recorder works ne with Chrome or Firefox on both Mac and PCs.

e Beware that the Quercus media recorder doesn’t allow pauses but you are able to

retake your videos as many times as you want.

f Quercus accepts media le uploads of size up to 500MB if you are uploading a le. The supported le types for playbacks are FLV, ASF, QT, MOV, MPG, MPEG, AVI, M4V, WMV, MP4, and 3GP. If you upload any other le types, the TAs may not be able to assign you a grade.

3   PART IV. Grading Scheme/Grading Rubric.

[Total Marks = 10+10+5=25] Report (submitted on Crowdmark) [Question box: Report]

DAE: Design and Analysis of Experiments. OS: Observational Studies

[Marks =10] 4.1 Description of the design including objectives of the study.

Grade (points)

Descriptions

Excellent (10)

DAE: Strong evidence of original thinking and a clear explanation of why and how they conducted the experiment.

OS: Strong evidence of original thinking and a clear explanation of why and why they carried out the study.

Good (8)

DAE: Grasped the basics of designing a factorial study; a good explanation of how and why they did the experiment.

OS: Grasped the basics of designing an observational study; a good explanation of why and how they did the study.

Adequate (6)

DAE: Understood the basics of designing a factorial study, but may not have designed a factorial experiment. Provided an adequate explanation of the design.

OS: Understood the basics of observational study, but may not have an observational study to work with. Provided an adequate explanation of the observational study.

Marginal (4)

DAE: Some evidence of understanding the basic design of a factorial study. Provided a poor explanation of their design.

OS: Some evidence of understanding the basic design of an observational study. Provided a poor explanation of their observational study.

Inadequate (2)

DAE: Little evidence of even a supercial understanding of a factorial design. Little expla- nation about how or why the design was chosen.

OS: Little evidence of even a supercial understanding of an observational study. Little explanation about how or why the observational study was chosen.

[Marks =10] 4.2 Analysis of the data (including methods, 2 pages maximum)

DAE: Include appropriate plots and calculations such as:  main effects and interactions; estimated variance of the effect (if replicated); confidence intervals for true values of effects (if replicated); Lenth plot; or half normal plot.

OS: Present screening of covariates, computation of propensity scores, approaches for the adjustment for treatment factor effect, you may use the R codes discussed during the lectures.

Grade (points)

Analysis of the data (DAE, OS)

Excellent (10)

Strong evidence of data analysis skills. Probably used R to do calculations and plots, but calculations and plots might also be done neatly by hand.

Good (8)

Good evidence of data analysis skills. Appropriate calculations were done,