The tool makes it relatively easy to get started. Everything is controlled from a single window that follows the analogy of an open Filofax, with two open pages for you to work on centre-screen and tabbed dividers left and right indicating the other program options. This is both convenient and efficient during the set-up stages, as you can drag and drop from one side to the other. Different pages allow you to choose the study you want to work on, the questions to go in the rows, the columns, filters and so on.

The product is a cross-tab tool, not a data mining or multivariate analytical tool, and it contains a wealth of options for setting up tables. You can define filters, group answers, combine and recode questions and put any number of items on the top, down the side or into the ‘level’ - the third dimension of the tables.

When you are ready, a run button generates the table and the display changes from personal organiser to something that looks disarmingly like Microsoft Excel. Indeed, it behaves very much like Excel in terms of applying stylistic changes. You can change the type size, font, colour, apply formatting rules, paste in graphics - but you just can’t change the numbers. You can also post the table to Excel proper, if the fairly standard range of presentation and charting options in mTABS don’t fulfil your needs. If you have applied a variable such as ‘month’ as a level, then a workbook of tabbed worksheets will be presented, each one relating to a separate month. It is easy to add an overall total sheet too.

Stepping through tabbed sheets like this makes year on year comparisons easy, and it is one of many thoughtful features for continuous research. An important one is its ability to join together different databases. A tracker could be delivered as a different database for each year, yet you could analyse it as if it were one large database. When you open more than one, you designate one to be the ‘primary study’. All the variables are then colour coded: red for a total mismatch, amber for slight differences you must resolve and green for a perfect match. As products and company names come and go, you can easily resolve those differences, and take the decisions yourself how to do this.

It is all pretty powerful stuff. Mean averages can be calculated on actual or midpoint values, and you can vary these. There is excellent support for hierarchical data through what are called ‘grouped’ variables, and you can flip between bases at one level or another with ease. Definitions, filters and even recodes can be saved, then reapplied later on the same database or another one.

There is a good range of significance testing, but the display was surprisingly crude: rather than converting results into significance levels (e.g. 95%), all you get is the raw standard deviation result. Furthermore, a constraint of the single spreadsheet workbook for output is that it only ever shows the current item you are on. There is no gallery of output previously produced as there is in, say, SPSS.

A rare MR customer is MORI, which uses mTAB on its Financial Services Omnibus, both in-house and to distribute databases to its subscribers. Technical Support Manager John Cogswell comments: “A lot of our clients see the value of being able to go along way beyond the basic cross tabs you can produce on paper. And using this, we can enrich our omnibus data with data from third parties, such as financial or geodemographic data. We went for this for its ease of use - it is incredibly user friendly with it’s drag and drop interface, and the Excel framework, means people feel they are familiar with it from the start.”
Among his favourite features are the traffic-light recoding of data between databases, and a feature called ‘top N’. “Essentially you sort the columns on a table independently of each other. If you are trying to highlight differences this has to be the easiest way to use it.”

MORI has been beta testing a new Java version, due out later this year. Cogswell approves of the move. “It will give clients the flexibility of accessing our data anywhere, and removes from them the responsibility for installing databases on their own network. Security procedures often mean this can take a long time: having the data outside, on the internet, can take that problem away.”