MCQ 1 Flashcards
(31 cards)
Psychologists tend to use parametric tests more often because…
They are less likely to lead to a Type II error.
We need to carry out statistical tests because…
Not all observed differences are meaningful.
We solve the problem of error variability by…
Using good design to reduce it, and statistics to estimate and account for what is left.
An experimenter has compared a sample mean to a population mean using a Z-Test and has calculated Z=3.8. What can we interpret from this result?
We can reject the null hypothesis as it is greater than the critical value of Z.
An experimenter has obtained a Z score of 1.76 for a single score drawn from a population with a SD of 5. What was the difference between the score and the population mean?
- 8
1. 76 X 5 = 8.8
As the sample size increases…
The standard error becomes smaller.
What is the difference between what standard deviation (SD) and standard error (SE) tell you?
SD tells us how far on average individual scores fall from the mean, whereas SE tells us how far on average the mean of samples of a particular size falls from the population mean.
An experimenter has carried out a statistical test and obtained a probability value of P
There is less than 5% chance that the result is due to error variability alone.
When comparing two means, the null hypothesis…
Always states that any differences between the two means are due to error variability.
Standard error for a repeated measures (related) t-test gives an estimate of..
Experimental error only.
Which of the following do we do when calculating SE for a between subjects (unrelated) t-test?
We add together the variance from both groups of participants.
Why does it matter how many people are in our sample when deciding whether our t-test is significant?
The number of people we test affects the shape of the t-distribution that we use to check for significance.
An experimenter has carried out a related t-test on data produced by a sample of 15 participants, and has obtained a ‘t’ value of 4. Which of the following is the most accurate AND MOST PRECISE way to report is result?
t(14) = 4, P
What is Type I Error?
A False Positive - says there is a significant effect, when really there isn’t.
What is meant by Type II Error?
A False Negative - says there is no significant effect when really there is.
Why do we use parametric tests instead of non parametric tests?
Parametric tests tend to be more powerful than non parametric tests.
What are the 3 assumptions for a parametric test?
1) Interval or ratio scale data.
2) Scores must be normally distributed.
3) Samples being compared must have similar variances.
True or False - Statistics eliminate error variability.
FALSE - Statistics allow us to estimate how much error variability there is so we can take it into account (THEY DO NOT ELIMINATE IT)
What do the results of statistical tests show?
The results of these tests (ratios) are the number of times greater our obtained differences are over those expected if the null hypothesis was true.
What does the SD tell us in the case of a single score?
The estimation of error variability in the set of scores.
What does the SD tell us?
SD tells us by how much on average scores in a sample differ from the mean of the sample (and each other) - It tells us how good the mean is a representation of the scores in the sample.
What is meant by the sampling distribution of the mean?
The distribution of all possible sample means of a sample of a specific size.
What does standard error tell us?
It tells us the average difference we would expect between the mean of a sample to population mean.
What does the p value mean?
It’s the probability of obtain the result we did when the null hypothesis is true. It is the probability that the obtained differences due to error variability.
