SOST70022

SOCIAL STATISTICS COURSE UNIT GUIDE 2022-2023 SOST70022 Longitudinal Data Analysis Semester: 2 Credits: 15 1 1. ESSENTIAL INFORMATION Lecturers Room Email Dr Alexandru Cernat G15 Humanities Bridgeford Street Building Alexandru Cernat: alexandru.cernat@manchester.ac.uk Teaching assistants Anthony: shingyan.kwong@postgrad.manchester.ac.uk Lena Andersen: lena.andersen@manchester.ac.uk by appointment via email; Lectures and labs will take place together Wednesdays 10:00-16:00 (see dates and topics bellow) Simon_6.004 Comp Cluster Assignment deadline for submission – 19 April (midnight) Assignment deadline for submission – 24 May (midnight) Office Hours with lecturer Times and Dates Lectures and labs: Location: Formative coursework Submission: Assessed coursework Submission: Assignments and Assessments • One 2500 word essay worth 100% of the total mark 2 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. Students are expected to be present in the computer lab unless they have received approval from the program director to study remotely for the second term. 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. 3 General Course Readings The following are the key texts for this course (see Blackboard weekly pages for detailed reading for each week): • Cernat, A. (in press). Longitudinal Data Analysis using R. LeanPub. • Singer, J., & Willett, J. (2003). Applied longitudinal data analysis: modeling change and event occurrence. Oxford University Press. (available online) • Newsom, J. T. (2015). Longitudinal Structural Equation Modeling: A Comprehensive Introduction. Routledge. The Course (week by week) Lecture and Lab 10:00 – 16:00 15 February 22 February 1 March 8 March 22 March Topic Introduction to longitudinal data Cross-lagged models Multilevel model of change Latent Growth model Survival analysis Location Simon_6.004 Comp Cluster Simon_6.004 Comp Cluster Simon_6.004 Comp Cluster Simon_6.004 Comp Cluster Simon_6.004 Comp Cluster 19 April* Formative assessment 24 May* Assignment deadline * deadlines are for the midnight of that day 4 • • • • • Lectures and Reading List Lecture 1: Introduction to longitudinal data Topics covered: Introduction to the concept of longitudinal data Data preparation and visualization Learning outcomes: Being able to prepare longitudinal data Being able to investigate transitions over time Being able to graph descriptive statistics Mandatory reading • Chapters 1, 2, 3, 4 in Cernat, A. (2023). Longitudinal Data Analysis using R, LeanPub. (pdf on Blackboard page) Additional reading • Chapters 1, 2 in Singer, J., & Willett, J. (2003). Applied longitudinal data analysis: modeling change and event occurrence. Oxford University Press. Link to book. 5 • • • • • • Topics covered: Introduction to lavaan Autoregressive models Cross-lagged models How to select between competing models Learning outcomes: Being able to estimate and interpret autoregressive and cross-lagged models Being able to use lavaan Mandatory reading Lecture: 2 Cross-lagged models • Chapters 5, 6, 7 in Cernat, A. (2023). Longitudinal Data Analysis using R, LeanPub. (pdf on Blackboard page) Additional reading • Chapters 1, 4, 5 in Newsom, J. T. (2015). Longitudinal Structural Equation Modeling: A Comprehensive Introduction. Routledge. Link to book 6 • • • • • Lecture 3: Multilevel model for change Topics covered: Introduction to the MLM for change How to estimate lme4 How to model and interpret the results of MLM using longitudinal data Learning outcomes: Being able to estimate a MLM for change Being able to interpret coefficients and choose between competing models Mandatory reading • Chapter 8 in Cernat, A. (2023). Longitudinal Data Analysis using R, LeanPub. (pdf on Blackboard page) Additional reading • Chapters 3, 4, 5, 6 in Singer, J., & Willett, J. (2003). Applied longitudinal data analysis: modeling change and event occurrence. Oxford University Press. Link to book. 7 • • • Lecture 4: Latent Growth Model Topics covered: The Latent Growth Model Learning outcomes: Being able to estimate a Latent Growth Model Interpret the results of LGM Mandatory reading • Chapter 9 in Cernat, A. (2023). Longitudinal Data Analysis using R, LeanPub. (pdf on Blackboard page) Additional reading • Chapters 7, 8 in Newsom, J. T. (2015). Longitudinal Structural Equation Modeling: A Comprehensive Introduction. Routledge. Link to book 8 • • • • • • Topics covered: Understand time to event data Discrete time event models Cox models Learning outcomes: Being able to estimate hazard models Being able to model survival/hazard functions Being able to model continuous time to event data Mandatory reading Lecture 5: Investigating event occurrence • Chapters 9, 10, 11, 13, 14 in Singer, J., & Willett, J. (2003). Applied longitudinal data analysis: modeling change and event occurrence. Oxford University Press. Link to book. Additional reading • Chapters 12 and 15 in Singer, J., & Willett, J. (2003). Applied longitudinal data analysis: modeling change and event occurrence. Oxford University Press. Link to book. 9 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 longitudinal data. This is expected to be in the form of a mini research paper. You are free to choose the research topic and dataset you are interested in. I recommend discussing these with the lecturers 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. The final submission should use real world social data. As such, you cannot use a teaching dataset or toy datasets (like those that come as part of computer 10 packages or teaching materials) or time series data (e.g., stocks, bitcoin, gdp, etc). Examples of suitable datasets are discussed in week 1. If you are unsure if the data you are planning to use is appropriate confirm with the lecturer in advance. The assignment should include as an appendix all the R code used to clean the data, create graphs and tables and run the analysis. This will not be considered towards the word count limit. Grading system for final essay: Introduction (20% of final grade) – 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 Data and methods (30% of final grade) – 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% of final grade) – 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% of final grade) – 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 11 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 Circumstances and extension requests 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 12 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 left hand 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. 13 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 may be 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 14 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’s guidelines 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/