In a simple regression analysis Seda finds a 95% confidence interval of [-1.4, -0.6] for the unstandardized slope. Is there a statistically significant association between the predictor and the outcome at alpha = .05? Why?
Like Salvatore S. Mangiafico, I am confused by what David L Morgan is saying. If "simple regression analysis" in the original question means ordinary least squares (OLS) simple linear regression, and if the 95% CI for the unstandardized slope is -1.4, -0.6, then the slope must be -1, the midpoint between those two limits. And as Salvatore suggested, the linear association between X and Y is statistically significant (at the .05 level), because the 95% CI excludes a value of 0. HTH.
p.s. - The wording in the original post (e.g., "Seda finds a 95% confidence interval of [-1.4, -0.6] ...") makes me suspect it is a homework problem, and that the value of the slope has not actually been given.
You did not say what the estimate was for the slope itself. If it is between [-1.4, -0.6], the result is non-significant at the .05 level. If the slope is outside that range, then it is significant at p
David L Morgan , maybe I'm not understanding something in the question. But if the confidence interval for the slope doesn't contain zero, doesn't this suggest that there's a significant relationship between the predictor and outcome variables? Is there something more complicated going on in the question?
Suppose the value was -2.0, that would be well outside the confidence interval, but if it were -1.0, it would be inside, and thus non-significant.
Salvatore S. Mangiafico, I see what you mean and I apologize. The basic idea is that the value of 0 is not within plus or minus 2 standard errors of the estimated value, so there must indeed be a significant negative effect.
Like Salvatore S. Mangiafico, I am confused by what David L Morgan is saying. If "simple regression analysis" in the original question means ordinary least squares (OLS) simple linear regression, and if the 95% CI for the unstandardized slope is -1.4, -0.6, then the slope must be -1, the midpoint between those two limits. And as Salvatore suggested, the linear association between X and Y is statistically significant (at the .05 level), because the 95% CI excludes a value of 0. HTH.
p.s. - The wording in the original post (e.g., "Seda finds a 95% confidence interval of [-1.4, -0.6] ...") makes me suspect it is a homework problem, and that the value of the slope has not actually been given.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Regression