Stats test 2
Terms in this set (67)
__ is a type of hypothesis that states the predicted effect does exist
__ is a t-test that establishes whether two means collected from two different samples differ significantly
__ is a type of sample variance that must be calculated during an independent t-test when the sample sizes are unequal
__ this type of test, aka a directional hypothesis, is used to determine if there is a significant relationship between variables in one specific direction
__ you do this to the null hypothesis when the sample mean is associated with low probability of occurrence when null hypothesis is true
__ test that tests the hypothesis that the variances in different groups are equal
__ type of hypothesis that states the predicted effect does not exist
__ a t-test that establishes whether two means collected from the same sample differ significantly
__ a t-test that establishes whether a sample mean differs significantly from a population value
__ this type of test aka a non directional hypothesis, is used to determine if there is a significant relationship between variables in any direction
__ type of error that occurs when we believe that there is no effect in the population, when in fact there is
__ type of error that occurs when we believe that there is a genuine effect in our population, when in fact there isn't
Formulas cheat sheet
If my experimental hypothesis were "Eating cheese before bed affects the number of nightmares you have," what would the null hypothesis be?
The number of nightmares you have is not affected by eating cheese before bed
If my null hypothesis is "Dutch people do not differ from English people in height" what is my alternative hypothesis?
All of the statements are plausible alternative hypotheses:
Dutch people are taller than English people.
English people are taller than Dutch people.
Dutch people differ in height from English people.
What does a significant test statistic tell us?
That the test statistic is larger than we would expect if there were no effect in the population, there is an important effect, and the null hypothesis is false
A Type 1 error occurs when:
We conclude that there is an affect in the population when in fact there is not.
A type 2 error occurs when:
We conclude that there is not an effect in the population when in fact there is.
"Children can learn a second language faster before the age of 7" is this statement:
A one tailed hypothesis
What is the alternative hypothesis for the following question: Does eating salmon make your skin glow?
People who eat salmon will have a more glowing complexion compared to those who don't.
What is the null hypothesis for the following question: Is there a relationship between heart rate and the number of cups of coffee drunk within the last 4 hours?
There will be no relationship between heart rate and the number of cups of coffee drunk within the last 4 hours.
Of what is the p probability?
P is the probability of observing a test statistic atleast as big as the one we have if there were no effect in the population (i.e., the null hypothesis were true)
Under a null hypothesis, a sample value yields a p-value of .015. Which of the following statements is true?
This finding is statistically significant at the .05 level of significance
What is a significance level?
A pre-set level of probability at which it will be accepted that results are due to chance or not
A researcher measured a group of people's physiological reactions while watching horror films and compared them to when watching comedy films. The resulting data were normally distributed. What test should be used to analyse the data?
Paired samples (dependent or related) t-test
Which of the following is an accurate description of the standard error?
It is the standard deviation of the sampling distribution of a statistic
The degrees of freedom for the paired samples t-test are:
Three types of t tests
1. Independent sample
2. Paired sample
3. One sample
To go from two tailed to one tailed...
Divide by 2
How is a t statistic calculated?
Mean difference/standard error of mean
When analysing the cancer survival data, a p-value set at .05 or .01 is most likely to protect against which type of error?
Type 1 error
Consider an independent t test, which of the following expressions best represents the degrees of freedom?
If a significance result of .000 is presented, how should this be reported?
p < .001
Steps to hypothesis testing
1. State the hypotheses (alternative vs null)
2. Set the criteria for a decision (.05, .01, etc)
3. Compute the test statistic (conduct the analysis)
4. Make a decision (supported or not supported)
Alternative hypothesis (H1)
That an effect will be present. What we usually think of when coming up with a prediction. Example: if you imagine eating chocolate you will eat less of it
Null hypothesis (H0)
Opposite of alternative hypothesis. Usually states that an effect is absent. Example: if you imagine eating chocolate you will eat the same amount as normal
__ is a mathematical formula that allows researchers to determine the likelihood of obtaining outcomes if the null hypothesis were true
Value of the test statistic is used to make a decision regarding the __
1. State the hypotheses
-Start by assuming null is true (just like defendant)
-But we think our hypothesis is true
-Conduct a study to find evidence that null isn't true
-Null and alternative hypotheses will involve a population (what were interested in generalizing to)
2. Set the criteria
-Significance level or level of significance
-Criterion of judgement upon which a decision is made
-Need enough evidence to show guilt beyond a reasonable doubt
-Behavioral research=5% usually
-Based on the probability of obtaining a statistic measured in a sample if the null hypothesis were true
3. Compute the test statistic
Value of the test statistic is used to make a decision regarding the null hypothesis
4. Make a decision
2 decisions researchers can make:
1. Reject the null hypothesis: sample mean is associated with low probability of occurrence when null hypothesis is true
2. Fail to reject the null hypothesis: sample mean is associated with high probability of occurrence when null hypothesis is true
How do I make that decision?
-Look at your p value. How likely it is to get a result like this if null is true. Compared to a level of significance (criteria)
-Is the p-value significant? Usually p<.05 in behavioral research then reject null
__ is rejecting a null hypothesis that is actually true
Type 1 error. False positive. You're found guilty of a crime you didn't commit
__ is not rejecting a null hypothesis that is actually false
Type 2 error. False negative. You're found guilty but not found guilty in trial. Told you don't have a disease when you really do.
If a test statistic has a p value <.05 or 5%, what is the decision for this hypothesis test?
Reject null hypothesis
One tailed hypothesis
Directional, hypothesis specifies some direction
Two tailed hypothesis
Non directional, hypothesis doesn't specify any direction
Type of statistic that is used in hypothesis testing. Inferential statistic used to determine # of SD's in a t-distribution that a sample mean deviates from the mean value or mean difference stated in null hypothesis
T statistic formula in words
t=(observed difference between sample means)-(expected difference between population means if null hypothesis is true)/(estimate of the standard error the difference between two sample means)
__ used to test hypotheses concerning single group mean selected from a population
One sample t test
__ used when there are 2 experimental conditions and different participants assigned to each condition
Independent t test
__ used when there are 2 experimental conditions and same participants took part in both conditions
Paired samples t test
What information do you need in order to find the critical t?
1. Two/one tailed
2. Degrees of freedom
3. Criterion (.05 or .01)
In one sample t test, if p<.05 it is ___
If obtained t is higher than the critical t then we __
Reject the null
3 steps for calculating unequal sample size
1. Pooled sample variance
2. Estimated sample error for the difference
3. T test formula
3 steps for calculating equal sample size
1. Sample variance
2. Estimated sample error for the difference
3. T test formula
Degrees of freedom for independent samples
__ assesses the assumption that the variances are equal
If null: p>.05 then...
Look at the 1st line (equal variances assumed), the variances are equal
If alternative: p<.05 then...
Look at the 2nd line (equal variances not assumed), the variances are not equal
2 methods for paired sample t test
1. Pre-post design
2. Within-subjects design
__ is the measure dependent variable before (pre) and after (post) a treatment
__ is when they observe participants across many treatments but not necessarily before and after treatment
The __ the value, the less likely a sample mean difference could occur if null hypothesis were true