forum: Quantitative Data Capture and Statistical using Access and SPSS

I'd be grateful for members' thoughts on the following reserach methodology problem:

Research overview

1. I am developing a new way to analyse historical technical drawings, so that they can be more appropriately made accessible, by archivists, for research use.

2. To do this, I intend to statistically sample large numbers of drawings from different collections – up to several thousand at least, in samples of c400.

3. Simplistically, the method will be to intellectually deconstruct individual drawings down into their single information elements – title, reference number, drafter, medium of the information carrier, etc etc.

4. For any one set of information element, I propose to record - does this element occur, yes or no. This will allow me to simply derive a percentage of yeses and nos for all drawings, at various levels of aggregation. I shall probably need to do more detailed statistical analyses as well.

5. Those elements that most commonly occur over all drawing types – more than 90% of the time, say, will be investigated in more detail, as they are the ones that archivists and researchers will most commonly see.

6. I am at the stage of having the first sample, and developing the tools to undertake the analysis.

ICT Methods Issues

I have three ICT Methods issues (I am quite IT literate):

1. Advice about the general appropriateness of my proposed IT approach, which is:

I need to statistically analyse the sample data, to extrapolate to larger populations. Potentially, shall also need to do more in-depth statistical analyses, depending upon what comes out of the research. For this, I am advised that I need to use SPSS – no problem there, I have started to learn it.

To capture the data, I am advised that I need something like would like to use MS Access data entry tables, which I shall have to construct from scratch. I shall determine their fields from the information elements that I find in the pilot – OK, I have used Access to a limited degree, and I can learn how to contruct data entry tables.

2. Assuming that that approach is satisfactory, what I am not sure about, before I start, is the interface between Access and SPSS, and how that will affect how I construct the data entry tables.

For example, for any one information attribute, there may be a number of possibilities. For information carrier, the options will be, paper, tracing paper, drafting film, polyester film, oil-impregnated paper etc etc. A drop-down list in the data entry table is obviously indicated, but how will that import as a yes – no data element in SPSS? Shall I have to enter yes – no in Access to every possibility, or could an Access import routine “automatically” make one option “yes” in SPSS, and all the others “no” ?

3. There will no doubt be other questions that I have not yet considered, and that is the nub of the issue – I am getting rather over-faced with the amount of information that is available, on the one hand, while not finding whether this sort of approach has been used before, so that I can apply it, and get on with the innovative part of developing the new analysis technique.

I understand that that is partially what research is about, but at the same time, as a mature career-break student from industry, I also appreciate that it’s better to apply available tools and solutions where possible!