This is a more theoretical question. I approached a term paper with the hypothesis that a decrease in a country's national GDP will lead to a decrease in the opposition's consensual parliamentary behaviour.
Now the issue I'm having is that when calculating estimated coefficients for this relationship, I plot GDPgrowth against Consensual voting. The regression method automatically calculates with an increase of unit in the independent variable, not a decrease. The results I got indicated that a one unit increase in GDP growth was associated with a ~-0.5 decrease in consensual voting behaviour - the opposite of what I predicted. This is not the issue to me though. The problem is, does the regression actually test my hypothesis? Because I hypothesised if a DECREASE in GDP Growth was associated with a DECREASE in the consensual voting behaviour, does that invalidate the regression model which calculates how an INCREASE in GDP Growth results in a change in consensual voting behavoiur? Or can I infer from the above results, that there exists a negative relationship between the variables where as my hypothesis predicts a positive relationship?