Which Method?

#1
Hi All,

I'm trying to find a suitable statisical method to analyse my data. I have data on what caused failiures for HGVs in an annual, arranged test vs a roadside inspection, and the number and rate of each failure. Each of these categories I have one value per year for six years. I have attached a sample graph to demonstrate what I mean.

I thought as first a paried sample t test, as I had "paired" data (same trucks, in two different scenarios), but I'm not trying to compare the means of the groups, I'd like to compare the differences over time. Same reason why I couldn't do an ANOVA.

Then I wondered about Chi test, by making one set of data the "null hypothesis" and comparing them both together, and using the categories of the failiures as discrete datsets, but I couldn't get my data to fit that either.

I've done many flow charts and questionaires asking "Which statisical anaylsis should I use?" and I've not managed to get the answer to work with my data. Please help!
 

Attachments