Update 22/10/16. The 2016 data is available here.
The full details of the methodology can be found in: J.Canning (2015) A new measurement and ranking system for the UK National Student Survey. Perspectives: Policy and Practice in Higher Education 19(2), pp.56-65. The publically available National Student Survey (NSS) data can be downloaded from the HEFCE website.
Calculating the Weighted Student Satisfaction Score (WSSS) and the Weighted Student Satisfaction Quotient (WSSQ)
Part 1 (calculating WSSS)
The main components of the model are the proportion (min, 0, max 1) who agreed or strongly agreed with each of the questions for a particular course, a weight for each of the questions (derived from Marsh and Cheng’s  factor analysis) and an adjustment for overall subject differences.
This overall score is then multiplied by 100 to avoid the overuse of decimal places. This calculation has been performed for all 4128 courses which appeared in the National Student Survey (2014).
Part 2: standardising the WSSS scores (WSSQ).
The resulting WSSS scores are then standardised to a mean of 100 and a standard deviation of 15. An average course will score 100.
|100||A course with a WSSQ of 100 would be average for 2014 NSS.|
|115||A course with a WSSQ of 115 would be exactly 1 standard deviation above the average 2014 NSS.|
|130||A course with a WSSQ of 130 would be exactly 2 standard deviations above the average 2014 NSS.|
|145||A course with a WSSQ of 130 would be exactly 3 standard deviations above the average 2014 NSS.|
|85||A course with a WSSQ of 85 would be exactly 1 standard deviation below the average 2014 NSS.|
|70||A course with a WSSQ of 70 would be exactly 2 standard deviations below the average 2014 NSS.|
|55||A course with a WSSQ of 55 would be exactly 3 standard deviations below the average 2014 NSS.|
All 5028 courses in rank order Download (.pdf)
Rank by subject (good for printing out in part or full) Download (.pdf)
Full dataset of WSSS and WSSQ for basic manipulation Download (.xlsx)
Full dataset with working out. This file is suitable for full manipulation/ experimentation. This file is in comma-separated values (.csv) format and can be opened with spreadsheet software (e.g. Excel or Calc) or imported into statistical analysis software. NB: The top row of the spreadsheet provides indication of the how each part of the sheet was calculated. .csv files are a text format so there is no formatting or calculations in the spreadsheet. Download .csv (24.1 MB). This spreadsheet contains 114,197 rows!
Wikipedia article on the .csv format.