Question | Answer |
What is the true slope? | The mean slope you would obtain if you repeated your analysis innumerable times using a series of similar samples. |
What 2 approaches are there to test the uncertainty around the true slope? | 1. Calculate a confidence interval around the slope. 2. Test the statistical significance of the slope by calculating a t ratio. |
Range of values around the slope with a given probability of including the true slope. | Confidence interval |
What does a 95% confidence interval say about the true slope? | There is a 95% probability that the true slope will fall within the confidence interval. |
What is the amount of uncertainty estimated by? | The standard error of the slope |
True/False: The null hypothesis is the proposition that the true slope is one. | False ... that the true slope is ZERO. |
What does a null hypothesis imply? (2 reasons) | (1) The independent variable has no effect on the dependent variable, and (2) the slope calculated in a sample is a result of nothing more than chance. |
What would support the null hypothesis? | If the slope is judged to occur frequently by chance. |
What is a research hypothesis? | A hypothesis that purports a relationship between the independent and dependent variable. |
What criterion constitutes "infrequently by chance"? What is it referred to as? | It should occur no more than 5% of the time by chance (referred to as alpha). |
When is a slope said to be statistically significant? | When its frequency of occurring by chance is less than or equal to 5% (alpha). |
Which alpha percentage gives more leniency to a statistically significant number: 10% or 1%? | 10% - to be considered statistically significant the slope should occur no more than 10% of the time by chance. |
In reference to the comparison between t(calc) and t(table), when is a slope significant? | When the absolute value of t(calc) is greater than or equal to the t(table). |
Correct positive | The slope is found to be statistically significant and, in reality, the true slope is not zero. |
Power | The probability of a correct positive decision (1 - beta). |
Correct negative | The slope is not statistically significant and, in reality, the true slope is zero. |
False positive | The slope is found to be statistically significant, but, in reality, the true slope is zero. |
A false positive is often called a Type One Error. What is the probability of making a type one error equal to? | It is equal to alpha. |
False negative | The slope is not statistically significant, but, in reality, the true slope is not zero. |
A false negative decision is called a Type Two Error. What is it equal to? | 1 minus power (remember: power refers to the probability of getting a correct positive) - termed beta |
True/False: The steeper the slope, the weaker the effect of the independent variable. | False |
True/False: The effect of an independent variable is not a function of sample size. | True |
Other things equal, how does a larger sample size affect the t ratio and probability? | As the sample size increases: - t ratio gets larger - probability gets smaller |
Give an example showing that the effect of an independent variable is not a function of sample size. | Scotch's effect on the central nervous system. |
Want to create your own Flashcards for free with GoConqr? Learn more.