EXISTING SITUATION - The MIS System

The MIS was in fact found to be two quite separate projects (each with a different UNFAO expert attached) with similar sampling, data collection and data presentation techniques, but with quite different data input and data processing methods.

People.

One of these projects had run in Dar es Salaam for the Dar Pugu market, with the supervisor being Jim Airey; the other had run in Moshi for the Moshi Weruweru market and the Arusha Themi market, with the supervisor being Martin Doran. Airey was active 1989 through August 1994, and Doran was active approx 1992 through December 1993. Both had statistical backgrounds, and both were working for UNFAO, with their immediate supervisor being based in Dar es Salaam. Airey operated from the Marketing Development Bureau in Dar, whereas Doran operated from the RDD offices in Moshi. To complicate matters, the Tanzanian counterpart to Airey, James Bokella, left for long-term study in the UK 7 days after I arrived on the Project (he has been replaced by Mr Kadigi). The Moshi and Arusha counterparts, Mr Swai, Mr David Kilimba, and Mr Ali Semajaila, are still on the Project.

Documentation.

There was little documentation on the computer systems, and also little documentation on the sampling, data collection, data input, data processing, and

data presentation. There were some sections of other documents which produced numbers and statistics from the MIS programs, e.g. as regards trends in yardage, sales, grade and price with seasonal and year-to-year changes, but the last of these had been done in 1992 (These are listed and reviewed in the appendix of visits and discussions). There was a 4 page xeroxed users' manual for the Dar es Salaam computer system, which covered the practicalities of operation only. No documentation has ever yet been found on the 'Marker' system which was used to smooth data for the Dar computer system.

Raw Data

The original data recording sheets were as far as can be determined all available for use - the Dar data (and some data for Korogwe) at the Dar MDB Office, and the Arusha and Moshi data at the Moshi RDD Office.

The processed weekly and monthly reports were also all available at the same locations.

Raw data existed for Dar for the period 1989 - 1994, with some gaps, and for Moshi and Arusha for the period 1992 - 1994, again with some gaps. Sometimes the gaps were due to rains when it was physically impossible to get to the market sites using motorbike.

Computer Data

The computer files with the computer raw data and the computer processed data were found in the Dar MDB Office and the Moshi RDD Office on computer hard discs and on computer floppy discs. These were relatively complete for the Moshi and Arusha data, but with a few gaps for the Dar data. The gaps in the Dar data may be due to the theft of a notebook computer in 1993 from the MDB Office, for which insufficient or no backup discs existed. If it is required now or in the future, and if it is considered necessary and economic, then the missing data could be reinput from the raw paper records. This applies also to some gaps in the Moshi and Arusha data.