Association / Inference

Question 1

z=0 and x=42. What is the mean?

Answer 1

A z of 0 indicates that, if the distribution is normally distributed you are AT the mean. So the mean is 42. And the median will likely be 42 as well.

Question 2

Are the results of a larger sample size or a smaller sample size more credible?

Answer 2

Given a random or stratified sample, larger sample sizes will tend to give better results.

Question 3

Is a significance level of .01 or of .05 more stringent?

Answer 3

.01 is more stringent - it means you have to have a bigger difference before you are willing to say there is a difference.

Question 4

If the significance level is .01 what is the probability of a Type I error?

Answer 4

.01 Alpha is the probability of a Type I error

Question 5

If the prediction equation is Y=3x - 6 is this a useful regression equation?

Answer 5

Well a correlation would tell us the utility of using the equation. And then we would have to decide if that ability to predict is useful for our purposes or not.

Question 6

The sample mean we obtain is significantly different from 0. Does this indicate that the population mean is different from 0?

Answer 6

It indicates that there is LIKELY some difference from 0 in the population.

Question 7

Is a significant difference an important difference? Why or why not?

Answer 7

A significant difference means that the difference (which may be so, so, so small) is real (ie. it is also the case in the population). Importance is never determined by a number - it is determined by whether for your purposes the difference is sufficient to do what you are trying to do.

Question 8

With the non-probability sampling methods you do not know the likelihood that any element of a population will be selected in a sample

Answer 8

The non-probability sampling methods do not give the likelihood (or probability) by definition.

Question 9

The set of methods using samples to estimate population parameters is _____

Answer 9

Statistical inference

Question 10

What is the unbiased estimator of the population mean?

Answer 10

A sample mean is the unbiased estimator of the population mean because the mean of all sample means will equal mu.

Question 11

What is the measure of variability most affected by extremes?

Answer 11

The range

Question 12

The level of significance (alpha) is the same as which type of error?

Answer 12

Type I error

Question 13

Convenience sampling is . . .

Answer 13

Non-probabilistic sampling - you don’t know the probability of any particular sample.

Question 14

A random sample of 121 bottles of cologne showed an average content of 4 ounces. It is
known that the standard deviation of the contents (i.e., of the population) is 0.22 ounces.
In this problem the 0.22 is

Answer 14

a population parameter

Question 15

\_\_\_\_\_\_\_\_\_\_ are used to infer that the results from a sample are reflective of the true population scores. 
  A) Descriptive statistics 
  B) Regression statistics 
  C) Correlated statistics 
  D) Inferential statistics

Answer 15

D) Inferential statistics

Question 16

A Type I error occurs when the null hypothesis is
A) rejected and the research hypothesis is actually false.
B) accepted but and research hypothesis is actually true.
C) rejected and null hypothesis is actually true.
D) accepted and null hypothesis is actually true.

Answer 16

C) the null hypothesis is rejected when it is really true

Question 17

If a mechanic looks at your car engine and says there is nothing wrong with it and your car breaks down when you leave the garage, what type of error did the mechanic make?

Answer 17

S/he failed to reject the (null) hypothesis that your car’s engine is equal to all other car engines. Really your car’s engine is different - it is not functional! So that is a Type II error.

Question 18

The probability of a Type II error is related to \_\_\_\_\_\_\_\_\_\_. 
  A) sample size 
  B) significance level (alpha) 
  C) effect size 
  D) All of the above.

Answer 18

B) signficance level (alpha)

Question 19

Which of the following statements is TRUE?
A) A very low significance level increases the chances of a Type I error.
B) If the effect size is small, a Type II error is unlikely.
C) When the null hypothesis is rejected, the population means are equal.
D) True differences are more likely to be detected if the sample size is large.

Answer 19

D) True differences are more likely to be detected if the sample size is large.

Question 20

If the null hypothesis was rejected and there was 1 chance out of 100 that the decision was wrong, what was the alpha level in the study? 
  A) .01 
  B) .10 
  C) .001 
  D) .100

Answer 20

A) .01