False - Choosing Random Individuals Who Pass by Yields A Random Sample False - Probability Predicts What Kind of Population

Probability --A system of definitions and operations pertaining to a sample space.
Sample space -- a set of points and can represent anything.

Eventan observable occurrence of a point. For any event A, 1 P(A) 0
False--Choosing random individuals who pass by yields a random sample
False--Probability predicts what kind of population sample is likely to be obtained.
Probability is equivalent to proportions.
Sampling error-- is synonymous with variability due to chance it indicates that
the value of the sample statistic (M) will probably be in error/deviate from the
parameter ().
Hypothesis test-- is a statistical method that uses sample data to evaluate a
hypothesis about a population. Accounts for random variability that might lead to
false rejection or retention of H0
Sampling distribution of the mean ---is nothing more than the distribution of
obtained sample means for all possible samples of a particular size (n) drawn from a
population.
Calc this with central limit theorem which is sigma M= sigma/sqrt of n
expected value of M is equal to
The distribution of sample means can be described in terms of shape
If the population from which samples are selected is normally distributed.
OR, if the sample size selected is large, around 30 or more.
M tells you how much difference to expect from one sample to the next
Small M: little variability between sample means. sample means cluster
together around M.
Large M: lots of variability between sample means. sample means are
spread around M.
2 factors influence M :
the size of the sample selected and the standard deviation of the
population from which the sample was selected
z-scores allow us to describe the location of a given sample mean relative to the
distribution of sample means
Z-test
1. State hypothesis H0: mui=the given mui
2. Define critical regions (smaller critical regions=lower alpha levels) and it The
cri-cal region defines unlikely values if the null hypothesis is true
3. Obtain random sample and do the z-test using population error
4. Retain or reject the null hypothesis
When rejecting remember that:
It simply means that your experiment (i.e. your particular sample) did
not provide enough evidence to reject H0.
It does not mean that you proved H0 to be correct.
Absence of proof is not proof of absence.
The level chosen for a hypothesis test = probability of a Type I error.
Type I error is inherent to hypothesis testing, but the researcher can control
the probability of the a Type I error by choosing the level.
Type II error occurs when the researcher fails to reject H0 when it really is false.
The probability of a Type II error is denoted by .
Criteria for Z-test
1. Random sampling and independent observation
2. Standard deviation isnt change by the treatmentie its the same as the
populations
3. Normally distributed sample.
How to write a concluding statement for 2 sample Z-test
There was no evidence that attendance in the English composition course
significantly changed performance on the writing sample, z = 1.5, p > .05
Power calc
1.
2.
3.
4.
Get critical boundary z-score

Find the M using the given sample mean mui
Now using NEW sample mean get the z-score
Get the probability which is the power
Beta= 1-power
Influences on Power
As effect size increases, power also increases.
Larger sample sizes produce greater power.

Reducing the alpha level (making the test more stringent) reduces power.
Using a one-tailed test increases power.
Why use the T-test instead of the Z-test?
The t statistic is used to test hypotheses about an unknown population mean
when the value of is also unknown.
T-test
n-1 = degrees of freedom. Degrees of freedom vary according to sample
size.
In the z-score test we know therefore no estimation therefore no degrees
of freedom.
1. State the null (Ho: mui=0) and select an alpha level.
2. Locate the critical region using the t distribution table and df.
3. Calculate the t test statistic.
4. Make a decision regarding H0.
Criteria for T-test
-Values in the sample are independent observations.
-The population sampled must be normal
Two sample t-test
Between-subjects design:
2 (or more) separate and independent samples (in the case of a t test:
independent samples t test).
Each sample contains a separate non-overlapping set of subjects.
We compare these two groups.
Experimental vs. control group.
Steps:
1. State Ho: mui 1 = mui 2
2. Do steps 2-4 from one sample t-test
Df= n1+n2-2
Must do pooled variance if n1 doesnt = n2
Look out for homogeneity of Variance
Must do the f max test which checks that the variances are similar
How to write a concluding statement:
Students who watched sesame street had an average high school grade high
school grade of M = 93 (SD = 4.71 ), while students who did not watch sesame
street has an average high school grade of M = 85 (SD = 4.22). There was a
sigificant difference in average high school grade between students who
watched sesame street and student who did not watch sesame street, t (18) = 4,
p < .05, d = 1.79.
***if you have one tailed instead of two you need to change the Ho and the tcritical
For example Ho= mui 1 <= mui2 if you were looking for an increase in
mui 1
Criteria
1. Random sampling
2. Population = normal for both groups
3. HOVmust have equal variances
Within-subjects design:
One sample from which we obtain data twice.
A repeated-measures design (in the case of a t test: a related samples t
test).
We compare the two samples obtained from the same group of subjects.
Weight before/after a diet program.
Learning performance on a task across two trial blocks.
Steps:
1. Find the difference between the two tests done on the same subjects
a. Difference score = D = X2 X1
2. State the Ho: mui D =0
3. Do steps 2-4 from the one sample t-test
****if one-tailed then the H0: D 0 or D 0
How to write concluding statement:

Changing the background color from white to red increased attractiveness
ratings of the women in the photographs by an average of M = 3.00 points with SD
= 1.50. The difference was significant, t (8) = 6.00, p < .01, d = 2.00
Criteria for within-sample/related sample T-test
Observations within each treatment condition must be independent.
Population distribution of difference scores must be normal.
This assumption is not a concern unless the sample size is small.
With relatively large samples (n > 30) this assumption can be
ignored.
Pros and Cons of the related samples t tests
Advantages of repeated measures design
Requires fewer subjects.
Able to study changes over time.
Reduces or eliminates influence of individual differences.
Disadvantages of repeated measures design
Factors besides treatment may cause subjects score to change.
Participation in first treatment may influence score in the second treatment
(order effects).
Order effects can be addressed by counterbalancing.

False - Choosing Random Individuals Who Pass by Yields A Random Sample False - Probability Predicts What Kind of Population

Uploaded by

Document Informationclick to expand document information

Copyright:

Available Formats

False - Choosing Random Individuals Who Pass by Yields A Random Sample False - Probability Predicts What Kind of Population

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

False - Choosing Random Individuals Who Pass by Yields A Random Sample False - Probability Predicts What Kind of Population

Uploaded by

Copyright:

Available Formats

Probability --A system of definitions and operations pertaining to a sample space.

Sample space -- a set of points and can represent anything.

Get critical boundary z-score

Larger sample sizes produce greater power.

How to write concluding statement:

You might also like