This paper argues in favour of moving beyond simple preference ranking in P/RRA as conventional methods produce limited data which is often misinterpreted. While ranking enables participants to define their own criteria for discriminating between items, it does not give an overall preference order because different items may have different weightings, so simple adding up would give misleading results. Asking participants to list items subjectively from best to worst overcomes this but still leaves the difficulty of interpreting the gap between ranks. Another alternative is to ask participants to give points to all the items being considered, so that simply adding up the scores allows different items to be compared. A number of ways to improve scoring are presented and illustrated using examples taken from an assessment of a food-for-work programme in Merti-Jeju district, Ethiopia. The study was carried out with women in villages to investigate food preferences, and help choose appropriate commodities for the programme. Techniques include qualified preference which compares one item to another and 'shopping' which simulates choices between many items with different market values. It is considered that these methods give better quantitative information and provide a useful way to explore real-world preferences in different circumstances.
Interesting for researchers and planners concerned with extracting quantitative information from RRA methods