Skip to content The Open University

Student and tutor module reviews

Linear statistical modelling

see module description

  • Points: 30
  • Code: M346
  • Level: 3
On this page

Student reviews

I very much enjoyed this module. I know some students felt there was too little mathematical content but I think that misses the point. If you are doing the maths & stats degree you need to know how to apply your mathematical knowledge and that's what this module is all about. Developing the skills to properly interpret data is deceptively challenging and it wasn't until I was into the revision phase of the module that I started to fully appreciate how difficult it can be. Good judgement comes with experience so I would recommend allowing plenty of time for practicing past papers. In summary a good, interesting module, and an essential one for anybody studying statistics.

John Schofield

Course starting: October 2017

Review posted: December 2018

I personally found the written material (in the form of multiple booklets) here not as well written as previous maths modules with the OU. At times the text felt too prose like without the succinct and reinforced 'idiots guide to' nature of other modules which I personally missed.

This is a very 'practical' course; I feel having a mix of theory and then practical would have been better (in say 2 module parts) than this in isolation and then the mathematical statistics module later. The exam was fair.

Course starting: October 2017

Review posted: September 2018

This course is essential for mathematics and statistics students, but also of interest for those reading social or natural sciences. I counted this module towards a BSc in Economics and Mathematical Sciences and found the content extremely useful for my econometrics project. Thoroughly recommended.

Course starting: October 2013

Review posted: July 2015

I enjoyed this course and I'm not sure how much I can add to the earlier reviews as many of the comments still apply even after a rewrite in 2009.

I would say it's a course that might generate a lot of moderate marks i.e. relatively easy to pass but hard to get really high marks. There was usually some need for judgement and some answers are not cut-and-dried, and there were usually a few tricky bits towards the end of the TMA questions. The exam has the dreaded essay which has been running for many years it seems and the solution to this would seem to be lots of practice of topics from past papers.

Some of the underlying Maths is reportedly now covered in M347 and past complaints about carrying one big book around don't apply as it is in 12 separate units now, and also online of course.

I thought the material developed quite well in step-by-step fashion but it was only at the end when revising and doing essays that many connections became clearer so if you can manage to get to this revision stage early (unlike me) that might help your enjoyment and understanding. My tutor was excellent so get to tutorials if you can.

Course starting: October 2012

Review posted: August 2013

This was a very demanding course but also very rewarding. The course itself appears very easy since most of the time one is just analysing data using the computer software supplied. There is little maths involved. The TMAs are straightforward and the amount of work required light.

However, once the units are completed then there is the examination which is another matter. I finished working through the units and had completed the TMAs by August and then turned my attention to revision. Thats when the course really comes to life and by working through past papers and producing my own mark schemes and sharing these with other students on the forum did I really get to grips with the aims of the course.

This was the best part where all of the ideas developed in the units were brought together. The exam was challeging but very fair if you prepare well. I would recommend this course if you wish to develop the skills of data analysis.

Martin James

Course starting: February 2011

Review posted: December 2011

M346 is a course written around the operation manual for genstat, a software course for examining Statistical Data. From March til August it is a fairly simple plod clicking on icons on a computer screen to find out how X affect Y. However the work load is light, and although the TMAs are heavy on ink, it is not difficult to score highly.

Oddly in September once revision starts you are not tied to the computer so you can focus on what is going on and the course becomes quite interesting. It is possible to revise 1 unit a night (there are 12) in an hour or so without the computer and things become quite clear.

Beware though the exam without a computer is a bit like a driving test without a car, and you need to plough through pages and pages of second hand computer data. Worse is to come with 25% of marks for an essay.

The course on the whole is interesting (ish), the work load is light, the TMAs are simple but it is not easy to get a good pass in the exam.

Sean Robinson

Course starting: February 2010

Review posted: December 2010

I would recommend this course to anyone. It is the course for anyone who wants to add the practical statistics tools to their kit. I was initially very apprehensive, the past student reviews particularly spooked me, but in hindsight there was nothing to worry about. I thought they were a bit harsh but it was pointed out to me that the course has been rewritten and this was the first year.

It is not easy, and it still requires almost constant access to the software, but I found it very rewarding.

Course starting: February 2009

Review posted: December 2009

This course takes work and patience to get through. The course textbook is poorly laid out in comparison to other commercially available statistics books and other OU materials.

In particular (as noted by other reviewers) useful material is contained within the answers to exercises at the back of the book not within the body of the chapter text. This gets frustrating and in the end I resorted to copying relevant sections and constructing a "proxy-textbook" that had all relevant material together in one place.

The course is also an applied course therefore the supporting mathematics is kept to a minimum which I did not like.

The final issue I came across was a lack of past paper answers which made exam preparation difficult but my tutor was extremely helpful in pointing me in the right direction on a number of questions.

The exam contains an essay question which will force you to think across techniques and chapters however the questions are generally split into small 3-5 mark sub-questions so it is not as daunting as first impressions would suggest. Thinking through the past paper essay questions helped to consolidate the course prior to the exam.

Overall I don't regret taking the course but I did find it a bit of a slog!

Course starting: February 2008

Review posted: December 2008

With course text, software and a computer you can complete this course. Having said that, you do need all three almost all of the time, so it is NOT a very portable course.

You should be prepared to spend your time on the course in front of a computer, working on exercises and examples, and referring back to the course text. Information sometimes appears in the solution to an exercise that has not been mentioned in the preceding text - I found I needed a marker in the chapter I was on and one in the corresponding solutions chapter.

The materials are good (easy to follow) and the software is OK to work with, even for a technophobe.

The exam paper looks a bit daunting, with large chunks of printout for you to interpret, and after a computer-led course it feels a bit strange to complete the exam without one.

Angela Mary Bernadette Lyon

Course starting: February 2007

Review posted: April 2008

I found this course to be the worst in terms of the understanding I gained and skills I learnt out of all the OU courses so far (8 past and present).

The approach is markedly different from the other courses I have done, and did not suit me at all. For example very few of the underlying concepts are gone into in enough depth to appreciate what's going on, and many times important concepts are introduced in exercises which might be OK the first time through but is a nightmare on consolidating or revising. Moreover, there are no tv programmmes, multimedia, or the little applications that the M248 course had produced specially to illustrate particular points.

It's not difficult because of the workload; in fact the workload is pretty light. It's difficult if anything because the workload isn't hard enough mathematically, so that you get a surface level of confidence about what's been presented but a nasty shock towards the end when you realise you don't have enough depth of understanding to answer the rather challenging essay question in the exam. Although in principle methods are favoured over mathematical theory, in practise much of the course and the effort you put in is oriented towards learning how to use the course software package. This isn't an easy package to use, and so much of the effort on methods ends being spent in simply learning how to implement the method in the software package. Often I just felt like I was following instructions of what to type in to the software, without any real understanding of why.

Because there's no calculus or linear algebra (in a level 3 mathematics course!) the maths is straightforward. In comparison to the other level 3 statistics courses, M343 (Applications of probability), M343 is a lot more mathematical than M346. I'm doing M343 this year and in hindsight I wished I might have tried M346 after M343, as it might have helped to get a greater intuition about the underlying mathematics, which as I said, are rarely explained in M346 itself.

The first TMA is essentially covering the same material as M248, so you can get a "free" good mark without having to learn anything new. Overall it's definitely less work than M248 but the exam is a lot harder: in my year very few of the students got 85-100%. I had to resort to doing endless past papers to do well in the exam. I always prefer to revise the material itself and get a decent understanding with which to approach an exam. With M346 for the first time I had to resort to a cookbook approach, which got me the mark I wanted. As I'm not studying with the OU as a soft option to get an easy degree, I have to say I would have preferred a more challenging course that would have led to a greater level of understanding of the subject area.

I only did M346 because I want the Maths&Stats degree, and this is the main reason for doing it that I can see. However it would be worth considering if you wanted to get some idea of regression without the underlying maths, or were going to work in a field where the course software was being used. Prior to M346, I did MST121 and MS221 and M248 and I was really impressed: total opposite from M346.

Course starting: February 2006

Review posted: March 2007

This was a very poor course.
The textbook is basically a manual for GENSTAT and the course seems to be spent pressing buttons and getting the right answers rather than understanding the underlying statistical concepts. During a couple of very useful revision tutorials and masses of past papers it became much more obvious what the course was trying to teach, and it turned out that the ideas were extremely interesting.

It was frustrating that it only came together at the end as there was little enjoyment to be had during the course. Judging by the number of people who have said that going through past papers is essential in order to pass I think many people must have had similar experiences to me.

The book is poor and this course really needs additional covering material that puts everything you are doing in the book into a wider context and gets you thinking away from the computer screen. There is rarely more than 2 pages of text between computer exercises so you are very much tied to the desk throughout the course.
Very disappointing after the other courses I have studied.

Course starting: February 2006

Review posted: February 2007

I am a zoological researcher, and this course was an excellent preparation for using the generalized linear modelling approach in my research.

I was familiar with some of the topics, such as multiple regression, but I still benefited from the rigour with which these topics were covered. I especially appreciated the emphasis on analysing contrasts, particularly useful for anyone wanting to test specific hypotheses.

My one disappointment is there could have been something on mixed models - perhaps even just a note at the end with some suggested reading. The same could be done for structural equation models or path analysis.

Overall, well worth doing.

William Hoppitt

Course starting: February 2006

Review posted: February 2007

I distinctly disliked this course. It has a meandering style, doesn't cover the necessary mathematics properly, and isn't even particularly practical in that several commonly used techniques in ANOVA aren't covered. The contrast in the teaching materials between this and M248 could not be starker.

Course starting: February 2006

Review posted: December 2006

A really good course! The software isn't really all that clunky; what it is, is designed for functionality; sure it could have more shiny fonts in the various sub-windows [that all achieve a specific purpose, providing a full range of GENeral STATistical tools] although it really doesn't hark back to DOS; the way it is designed grows on you as you notice how it achieves its purpose.

The course itself is based on a well-written book; after familiarising yourself with the various [some quite advanced] notions involved, it guides you to learning to describe what you perceive in approachable language.

There is a good linkage between the various parts of the course as it evolves, so that after studying it the knowledge stays in the mind; it's not 'bitty' as certain other Level 3 Maths courses may be. I hope it's as enjoyable for you as it was for me!

Mark Benjamin

Course starting: February 2006

Review posted: December 2006

As has been already posted, this is very much a methods course, to help the prospective student to both analyse and interpret real life data.

Coming from a pure mathematician's perspective, I found this course enjoyable but rather difficult in knowing what exactly was required in the TMAs and the exam (both of which include writing an essay). In order to obtain good pass marks and an understanding of the material, a considerably amount of time is needed sitting in front of a computer working through the relevant exercises.

This module is completely different from it's counterpart (M343), is a relatively easy Level 3 course, and would certainly suit those who are required to do linear statistical modelling in the real world.

Course starting: February 2005

Review posted: July 2006

Please note

Each of the views expressed above is an individual's very particular response, largely unedited, and should be viewed with that in mind. Since modules are subject to regular updating, some of the issues identified may have already been addressed. In some instances the faculty may have provided a response to a comment. If you have a query about a particular module, please contact your Regional Centre.

Enter a module code to find a review

To send us reviews on modules you have studied with us, please click the sign in button below.