That P-value isn't small enough to reject the null hypothesis.
Can you throw up some of the original data?
I believe the p value for linear regression is: if you decide that Political view is linearly dependent on gender then the p value is the chance that they actually aren't linearly dependant. 0.266 is normally considered a large chance of incorrectly assuming a linear relationship so you would need to do more sampling.
A linear regression isn't really suitable here because gender only takes two values, 0 and 1.