辅导 SOST70022 Longitudinal Data Analysis 2023-2024讲解 Python程序

SOCIAL STATISTICS COURSE UNIT GUIDE 2023-2024

SOST70022 Longitudinal Data Analysis

2. COURSE CONTENT

Aim

To provide students with an understanding of different longitudinal designs and the skills needed to conduct appropriate analyses using longitudinal data. Methods covered include the multilevel model for change and models for investigating event occurrence over time.

Teaching Methods

The course will be divided in five topics. Each topic will be covered in a week and will have:

1. Readings. Each week you will have some mandatory reading as well as

some optional material to read. You are expected to do the weekly reading in order to fully understand the topics covered.

2. Lecture and practical. Each week will include a full day of teaching that combines a lecture with a hands on practical and a discussion of the solution. Studentsareexpectedtobepresentinthecomputerlab.

Objectives

• To gain competence in the concepts, designs and terms of longitudinal research;

• To be able to apply a range of different methods for longitudinal data analysis;

• To have a general understanding of how each method represents different kinds of longitudinal processes;

• To be able to choose a design, a plausible model and an appropriate method of analysis for a range of research questions.

Course

The UK is fortunate in having a rich and growing store of longitudinal studies for researchers to analyse. The course will introduce students to the methodological and statistical skills that will enable them to address questions about the measurement and explanation of change.

General Course Readings

The following are the key texts for this course (see Blackboard weekly pages for detailed reading for each week):

•   Cernat, A. (2023). Longitudinal Data Analysis using R. LeanPub. (free access

through Blackboard).

•  Singer, J., & Willett, J. (2003). Applied longitudinal data analysis: modeling

change and event occurrence. Oxford University Press. (available through university library)

•  Newsom, J. T. (2015). Longitudinal Structural Equation Modeling: A Comprehensive Introduction. Routledge. (available through university library)

The Course (week by week)

* deadlines are for the midnight of that day

The content (reading, practical, data, etc.) will be released on Blackboard a week in advance of the lecture.

Lectures and Reading List

Lecture 1: Introduction to longitudinal data

Topicscovered:

•   Introduction to the concept of longitudinal data

•   Data preparation and visualization

Learningoutcomes:

•   Being able to prepare longitudinal data

•   Being able to investigate transitions overtime

•   Being to describe longitudinal data using descriptive statistics and graphs

Mandatoryreading

•   Chapters 1, 2, 3, 4 in Cernat, A. (2023). Longitudinal Data Analysis using R,

LeanPub. (pdf on Blackboardpage)

Additionalreading

•   Chapters 1, 2 in Singer,J., & Willett,J. (2003). Applied longitudinal data analysis: modeling change and event occurrence. Oxford University Press.Linktobook.

Lecture: 2 Cross-lagged models

Topicscovered:

•   Introduction to Structural Equation Modelling (SEM)

•   Using lavaan to run SEM

•   Autoregressive models

•   Cross-lagged models

•   How to select between competing models

Learningoutcomes:

•   Being able to estimate and interpret autoregressive and cross-lagged models

•   Being able to use lavaan

Mandatoryreading

•   Chapters 5, 6, 7 in Cernat, A. (2023). Longitudinal Data Analysis using R,

LeanPub. (pdf on Blackboardpage)

Additionalreading

•   Chapters 1, 4, 5 in Newsom,J. T. (2015). Longitudinal Structural Equation Modeling: A Comprehensive Introduction. Routledge.Linktobook

Lecture 3: Multilevel model for change

Topicscovered:

•   Introduction to fixed effects and random effects

•   Introduction to the multilevel model for change

•   Estimating the multilevel model for change using lme4

•   How to model and interpret the results of MLM using longitudinal data

Learningoutcomes:

•   Being able to estimate a MLM for change

•   Being able to interpret coefficients and choose between competing models

Mandatoryreading

•   Chapter 8 in Cernat, A. (2023). Longitudinal Data Analysis using R, LeanPub. (pdf

on Blackboard page)

Additionalreading

•   Chapters 3, 4, 5, 6 in Singer, J., & Willett,J. (2003). Applied longitudinal data

analysis: modeling change and event occurrence. Oxford University Press.Linktobook.

Lecture 4: Latent Growth Model

Topicscovered:

•   The Latent Growth Model

•   Using lavaan to run Latent Growth Models

Learningoutcomes:

•   Being able to estimate a Latent Growth Model

•   Interpret the results of Latent Growth Model

Mandatoryreading

•   Chapter 9 in Cernat, A. (2023). Longitudinal Data Analysis using R, LeanPub. (pdf

on Blackboard page)

Additionalreading

•   Chapters 7, 8 in Newsom,J. T. (2015). Longitudinal Structural Equation Modeling: A Comprehensive Introduction. Routledge.Linktobook

Lecture 5: Investigating event occurrence

Topicscovered:

•   Understand time to event data

•   Discrete time event models

•   Cox models

Learningoutcomes:

•   Being able to estimate hazard models

•   Being able to model survival/hazard functions

•   Being able to model continuous time to event data

Mandatoryreading

•   Chapters 9, 10, 11, 13, 14 in Singer, J., & Willett,J. (2003). Applied longitudinal

data analysis: modeling change and event occurrence. Oxford University Press.Linkto book.

Additionalreading

•   Chapters 12 and 15 in Singer, J., & Willett, J. (2003). Applied longitudinal data

analysis: modeling change and event occurrence. Oxford University Press.Linktobook.

3. ASSIGNMENTS AND ASSESSMENTS

The assessment for this course will evaluate your ability to work independently and apply what you have learned to real life situations. As such, the final mark will be based on an essay where you answer a substantive research question using the English Longitudinal Study of Ageing (data access:

https://beta.ukdataservice.ac.uk/datacatalogue/series/series?id=200011). This is expected to be in the form. of a research paper.

You are free to choose the research topic, waves and statistical models you want as long as you meet the requirements below. I recommend discussing these with the lecturer and TAs in the class or during office hours.

In the first assignment (formative, non-assessed) you will be expected to present an introduction and the data and methods that you will use  (basically the first part of your research paper). I expect this will have a maximum 1500 words (including references) and will include the following aspects:

- context and why the topic is important (with appropriate references)

- research question(s) /hypotheses

- explain how you will answer the question and why longitudinal analysis is important

- present the data and the variables that you are going to use

- present descriptive statistics  (in tables/graphs) with a  special emphasis on the longitudinal aspects (eg. transition matrices, plots of trends, etc.)

- present the methods you aim to use  (preferably also  the  specific models  and sequence)

Assessed Coursework Details

Based on this work and the feedback received you are expected to develop a 2500 (excluding references, tables and figures) word research paper that will be assessed. The new paper will also have to include the statistical analysis and the conclusions (in addition to the improved sections submitted before).

Presentation is important! The writing style, referencing, tables/figures (and their titles) will also be taken into account. Within each section 10% of the grade will be based on this. Use the readings as guides for writing and presentation.

Minimum requirements for the assignment

Failure to meet all the requirements below will lead to automatic failure of the course:

1.  You must use at least four waves of the English Longitudinal Study of Ageing (data access:

https://beta.ukdataservice.ac.uk/datacatalogue/series/series?id=200011).

2.  You must use at least one of the following statistical models: cross-

lagged model, multilevel model for change, latent growth model or survival model. It is ok if you use just one type of analysis but I encourage you to have multiple models (e.g., you start with a simple model and build a more complex model, or have multiple models to test different research questions).

3.  The assignment should include an appendix with all the R code used to clean the data, create graphs and tables and run the analysis. Tbj∫ mjll not be  con∫jdeTed tomaTd∫ tbemoTd coMnt ljmjt.

Grading system for final essay

Introduction(20%offinalgrade)

-   Include the research question(s) with context why it is important (with appropriate references)

-   Explain how you will answer the question and why longitudinal analysis is important

Dataandmethods(30%offinalgrade)

-   Present the data and the variables that you are going to use

-   Present descriptive statistics

-   Explain what models (and their sequence) you are going to use

Analysis(40%offinalgrade)

-   Present the different models and interpret them

-   If choosing between multiple models motivate your decision

-   Investigate/discuss the assumptions of the model chosen

Conclusions(10%offinalgrade)

-   Restate the research question(s) and show how you answered it

-   Explain how longitudinal data analysis helped answer the question

-   Present limitations of the study

Coursework Submission

Coursework must be typed, double-spaced in a reasonable font (eg. 12 point in Times New Roman or Arial).

Essays should be submitted online via Turnitin by midnight on the deadline day given on p.2 above unless given course specific instructions by email. Ensure you have familiarised yourself with the system and give yourself plenty of time for

submission as technology problems will not be an acceptable reason for late or non-submission of work. If you have serious problems submitting on the day please contact the SoSS Postgraduate Office. When you have successfully

submitted your essay you will be able to download and print a receipt. You must

keep a copy of your submission receipt until all work on this course is complete and you have received your final grades.

Note that our online submission system includes Turnitin plagiarism detection

software. Be sure that you fully understand what plagiarism is; links for further

details are included in section 6 below. If, after reading the guidance, you are at all unsure about what counts as plagiarism then you should contact your Academic    Advisor to discuss it.

If your essay is submitted late your grade will be reduced by 10 marks per day for 5 days, after which it will receive a mark of zero. For clarity a ‘day’ is 24 hours,

beginning immediately after the published deadline. *Deadlines will be strictly

enforced in all cases*. The mark published through Turnitin will show your mark *before* the late penalty is applied. The final mark, with the late penalty applied,  will be recorded on the student system and used to calculate your overall course   unit mark.

Mitigating Circumstancesandextensionrequests

If you think that your performance or academic progress is likely to be affected by

your circumstances or that you may not be able to hand in your

assignment/dissertation by the deadline, you may submit a Mitigating  Circumstances form/extension request form, with relevant supporting

documentation, for consideration by the Mitigating Circumstances Committee and Board of Examiners.

The nature of the supporting documentation required will vary according to the    nature of the circumstances, but it must be sufficiently independent and robust to

confirm the veracity of the case you are making. Please note that it is your

responsibility as the student to submit a request for consideration of mitigating

circumstances by the published deadlines. You should not wait until your results

are issued or the deadline for the submission of your work to have passed to apply for mitigating circumstances as cases will not be accepted retrospectively.

4. FEEDBACK

All Social Statistics courses include both formative feedback – which lets you know

how you’re getting on and what you could do to improve – and summative

feedback – which gives you a mark for your assessed work. This course uses the following mechanisms for feedback:

•   Informal verbal feedback will be given during lectures and tutorials for   individual and group work. (You’ll need to contribute regularly to group discussions to make the best use of this.)

•   Written formative feedback will be given on your non-assessed assignment and made available via email.

•   Written summative feedback will be given on your assessed coursework, available via the Turnitin/GradeMark on the Blackboard system.

•   Exam results are published only as a grade. If you wish to discuss your exam performance with your lecturer please book an office hour slot by email and let your lecturer know in advance that this is what you want to do.

Your Feedback to Us

We’re continually working to improve our teaching practices – for that we need  your feedback. Towards the end of the semester you’ll be asked to fill out a Unit Survey for each of your modules – please do! The survey is designed to be very   short and easy to fill out but the results are really valuable for our monitoring of teaching quality. We want to hear from you whether your opinion on the course was good, bad or indifferent.

All of your Unit Surveys are available via Blackboard – simply go to ‘Unit Evaluation’ on the lefthand menu of the Blackboard website to begin.

Alternatively, you can download a smartphone app called EvaluationKit to fill out Unit Surveys for all of your course units.

5. YOUR COMMITMENT

Study Schedule

You are expected to:

•   Attend the live sessions;

•   To read all mandatory readings before each class;

•   You are strongly encouraged to read the additional reading;

•   Submit the formative assignment;

•   Submit the final assignment.

Attendance

You are expected to attend all live sessions that are part of your programme. It is also expected that you arrive on time.

Email and Blackboard

Your commitment is also to check your University email and Blackboard at least every other day in order to make sure that you are informed of any communications from tutors or administrative staff. These might, for example, concern important meetings with staff, changes of room; notification of course options registration, or course-relevant information from your lecturer. Being unaware of arrangements because you have not checked your email or Blackboard is not an acceptable excuse.

6. REFERENCING & PLAGIARISM

The lack of a proper bibliography and appropriate reference in assessed essay will potentially greatly affect the mark for the work and maybe considered plagiarism, which is a serious offence.

All essays must employ the scholarly apparatus of references and a bibliography. There are different acceptable referencing styles.  In Social Statistics we recommend use of the Harvard system of referencing, which is described in detail here:http://subjects.library.manchester.ac.uk/referencing-harvard

In short, Harvard referencing means that you refer to the author and date of publication in brackets within the text, wherever you are referring to the ideas of another writer. Where you quote an author you must always include quotation marks and a page number in the reference.

All essays must include a References List which lists your sources in alphabetical    order by author's surname. This should include all (and only) the sources you have directly referenced in the text. Whatever your source is, you need to provide a full  set of publication details as described in the guide linked above. All academic texts you read will include bibliographies and these should give you plenty of examples  of what information to include.

Plagiarism

The University defines plagiarism as ‘presenting the ideas, work or words of other   people without proper, clear and unambiguous acknowledgement.’ It is an example of academic malpractice and can lead to very serious penalties up to exclusion

from the University. You should read the University’sguidelines here:

http://documents.manchester.ac.uk/display.aspx?DocID=2870

There is additional useful guidance on plagiarism and referencing in the Crucial Guide:

http://www.studentnet.manchester.ac.uk/crucial-guide/academic- life/support/referencing-and-plagiarism/