which statistical test to use?

Question

conducted an experiment with:
independent variable - presence of a sign (nominal value)
dependent variable - instances of illegal parking (nominal value)

we had readings over 3 weeks (pre-test week without stimuli, treament week with stimuli, post-test without stimuli) at 2 carparks (one control, one treatment)

which statistical test should I use?
if chi-square, which numbers should I compare (week? carpark? illegal? legal?)

are there any websites you can direct me to since I don't have access to a statistician? most websites on chi-square seem to cover comparing of one IV (sign or no sign) with one DV only (no. of illegal or legal parkings). how do I add in the factor of control and treatment group?

design of the experiment:
http://img144.imageshack.us/img144/310/image1cx0.gif

Answer 1

You haven't defined stimuli, so I'm going to ignore it and I'm not sure it's important.
Set up a 2x2 table. Columns (j) would be car park sign vs car park no sign
Rows(i) would be # legal parked vs # illegal parked.
Fill in the numbers. Find the totals for each row and column.
Find the grand total.
Find the expected values, Eij, for each cell,
Eij = (ith row total)x(jth column total)/Grand total
Find the chi-square value, X²ij, for each cell.
X²ij = (Oij - Eij)²/Eij Where Oij is the observed value (your cell data)
Find the total X² value for all the cells.
Find the critical chi-square value from a chi-square table for your α risk (usually α = 0.05) and 1 degree of freedom, 3.841, for example.
If your calculated total X² value is greater than the critical X² value, then the hypothesis of no relationship cannot be accepted and conclude that signs discourage illegal parking.

Answer 2

With a well-conducted chi-square test, you can compare them all (legal vs illegal within week within carpark). Check with a statistician.

Answer 3

the respond on your question relies upon on the (a) precise speculation (study question), (b) the form in which you gathered your information, and (c) the dimensions equipment. in case you in basic terms have score scales for the two self-efficacy and relatives verbal replace form, i could advise an ordinary correlation (you haven't any longer have been given any experimental manipulation. devoid of understanding the character of the score scales, i'm no longer able to be confident that's the suitable correlational attempt, yet while the scales have some good psychometric values, you're able to pick Pearson's r. if your scales are susceptible and bring about some restrict of selection, you're able to be conservative and attempt Spearman's rho.