Quick Review: Test 2 Flashcards
What is the sampling distribution?
The distribution of a statistic over repeated sampling
What is sampling error?
the variability of sample estimates of some statistic such as the mean
What is the standard error of the mean?
the standard deviation of the sampling distribution of the mean
What do we mean by hypotheses testing?
We are referring to testing some hypothesis about the relationships between population parameters
What is a research hypothesis?
this is the hypothesis that there is a true difference or a true relationships between the variables
What is the null hypothesis?
the hypotheses that is apparent difference difference or relationship is due to change
- Saying things dont affect eachother
example: going through cognitive therapy does not affect an anorexic girls weight
Why do we test null?
the alternative hypothesis is to vague, where as the null hypotheses is specific and if we can reject it, we can argue that the alternative hypothesis is true
what is another term for rejection level?
the significance level
What is a type II error?
rejecting the null hypothesis, when the null hypothesis was actually true (false negative)
What is a critical value?
the value of the test statistic beyond which you reject the null hypotheseis
We often reject the null if the conditonal probabibility of the data, given tthat the null is tue, is less then .05? what symbol do we use to represent this?
alpha
if we reject the null only if the difference betweel the groups is POSITVE, we are using whwat kind of tailed test?
1 tailed
-because we are explaining one specific direction
List three things the central limit theorum tells us:
- the mean equals the population mean
- the standard deviation (standard error) equals the population standard deviation divde by the square root of n
- The distrubition approaches normal as N increases
why do we care about the standard error of a statistic?
it tells us how variable that statistic is over repeated sampling
What does formula for t differ from the formula for z ?
we replace Z raw score of (x), into t scores with a statistic of a sample (Mean)- and we replace population standard deviation, with the standard error of the mean
Why is the sampling distribution of the variance relevant for t tests?
the sampling distribution is positively skewed, especially for small samples, so any particular sample standard deviation is more likely to underestimate the population standard deviation, then over estimate it
when we are dealing one set of scores, what will the degree of freedom be?
n-1
name three things that affect the size of the t we calculate
the size of the difference, the size of the variance and the sample size
what do we mean when we speak of an effect size measure?
we refer to some measure that tells about how large a difference is, in a meaningful metric rather then weather or not its statisticaly signficant
what is the confidence interval
we are speaking about an interval that is calculated in such a way that it has a partciular probability (95%), of including the true population value of a parameter
What does cohens effect measure?
(d)
it is a measure of how far the sample mean differs from a population mean when expressed in terms of standard deviations
what is sampling distribution of t?
this is the distribution that the t statistic would take on over repeated sample if the null hypotheses was true
What do we mean by matched samples?
the observations come in pairs, such that the two items in the first row of data come from the same person, or are related in some way
the major advantage to matched samples are”
allow us to remove extraneous variance before calculating t
what is the usual null hypothees with a matched sample?
the null hypotheses is that the population mean difference from one measurement to aother, on the same person is 0
THERE IS NO RELATIONSHIP
give two advantages of a matched samples study over a study with independent group
matched samples control for individual differences, they reduce the influence of extraneous variables
what does the value of d= .95 tell you?
that there was a difference of .95 standard deviation units between the two means
whats the difference in standard deviation used to calculate (d) and the standarddeviation used to caluclate the confidence interval?
In the former we use the standarddeviation to pretest means, while in the latter we used the standard deviation of different scores
What is a carry over effect?
a situation in which something about the first measurement influences the second- for example: learning one task may interfere with the learning of a second task
Give two reasons why we would run an expriement with independent groups of scores:
sometimes we cannot measure the same object under different conditions (boys vs girls), and sometimes the effect of the first treatment will to strongly influence the effect of the second treatment
what is the difference between N and n
N: the total sample size
n: sample size in an indivdual group
what do we mean by pooling variances?
we take an average of the two variances, with the variance of each sample weighted by the size of the sample on which it was based
-we average it out, because one sample size was bigger then the other one
What is meant by the assumption of homogenity variance?
the assumption that our two samples came from a population with the same variance, regardless of the values of the population means
What do we mean when we say that the two effects are confounded?
we cannot cleanly tell the difference between the two groups.
what do we mean by a bi-directional hypotheses?
this is really another way to say TWO TAILED
how does calculation of confidence limits on the difference between two means differ from the calulcation when we have one mean?
the only difference is that for the former we use the difference between the means and the standard error of mean differences
what are error bars?
lines drawn on a bar chart showing soemthing about variability means- often represent one SD above and below the mean, but under other units also
