Oliver, 2010 July. Define target audience Decide on stastistics Make Data Requests Chase data requests Collate Data Check the data Run analysis.

Oliver, 2010 July

Define target audience Decide on stastistics Make Data Requests Chase data requests Collate Data Check the data Run analysis Produce report Deliver 1st draft of report

Internal to Southampton Solent University

External

Dean and Assistant Dean

Subject lecturers

Public relations and Marketing

Mode, Median, Mean

Discreet or continuous variables

Normal distribution

Chi squared, null hypothesis

Discuss suitable format for the reports with faculty managers and course leaders

Which statistics would be appropriate ?

Drop out rate

Degree results 1sts, 2.1, 2.2, 3rds

Masters results : number of distinctions

Vocational courses

Request the data extract or access to database from IT

Construct one or more queries. Maybe link data from different tables / systems.

  Manipulate the data to obtain a data set

suitable to run the statistics on.

Use a stats package or do by hand or using Excel functions.

Use a student id rather than a name, maybe generate a new key field

So that students can't be individually identified by the student id

If the data is grouped need an audit trail back to starting point

Show initial reports to faculty managers and course leaders and request feedback, customisation

Add / Remove columns, further analysis

Are all the sections of the data complete ?

Are they consistent, do parts of the data contradict each other ?

  How are nulls treated, no data, N/A, no

shows for exam

  Comma delimited, xml, flat file,

excel, .mdb, blogs, Wikis  Research : Wikipaedia versus journals

Tools for mapping data sets, eg Apatar forge

Co-worker, line manager

Sanity check, totals etc.

Agile techniques, code peer review

Investigate anomalies

Hard copy, Access report, email, web page,

Interactive web page linked to data, mySql, PHP, Java

   Confidence limits 95%

For subject lecturers  5 Technology Subject Groups:

Communications Engineering, Computing, Design, Engineering and Construction, Maritime and Coastal Studies  

Faculty as a whole

Architecture Business IT Civil Engineering Computer Networking Computing Construction Electrical/Electronic Engineering Engineering (General) Interior Design Mechanical Engineering Media Technology Music and Music Technology Outside Broadcasting Surveying

What are the key performance indicators ?

Career outcomes, did the course help job prospects ?

Meeting learning objectives 360 students, lecturers, exam boards, employers

Qualitative data - stories of students

Quantitative data - pass rate, number of distinctions

  Complaints - modular choice, overlap  Level - too technical, not technical enough  Drop out rates

Raising status of the organisation,

Helping to meet social or research objectives

  Helping to meet staff development

objectives   Time spent online to student portal ?  Amount of portal interaction between

students and teachers ?

Social presence, tweets, webcam, video phone, skype

Mandatory number of postings per week ?

In online coursed, terms such as communities of inquiry, communities of learners, and knowledge-building communities have evolved.

   Improvement from knowledge / educational level at

course entry  Endowments and annual giving from past students

Define target audience Decide on stastitistics Make Data Requests Chase data requests Collate Data Check the data Run analysis Produce report Deliver 1st draft of report

Any questions welcome

Thanks Olly

http://obromley.deoss.net/web/index/course_performance.htm