Quantitative Methods in Psychology - Part I
Any measurement made on the entire population is called a(n)
parameter
Statistic is to sample as __________ is to population
parameter
The total number of observations sharing at least one trait in common is called the
population
A method of sampling in which every observation in the entire population has an equal chance of being selected is called
random sampling
When the calculated value of t is equal to or greater than the table value of t for a given number of degrees of freedom, the researcher should __________ the null hypothesis.
reject
Every point on the abscissa of a sampling distribution of means represents a(n)
sample mean
The difference between the sample mean and the population mean (M-μ) is called
sampling error
A point estimate is a sample value that is said to provide the "best" estimate of a single parameter value
true
If there is no difference (no relationship, no zero) between fo and fe, the value for chi square is
0
Because of sampling error, a point estimate can never be guaranteed to produce the true population parameter
True
The probability of committing the Type 1 error, or of being wrong when the null hypothesis is rejected
alpha level
μ1≠ μ2
alternative hypothesis
Central Limit Theorem
if we take a successive number of random samples, the plot will approximate a normal curve.
When the null hypothesis is rejected, even though it should not have been rejected, the __________ error has been committed
type I
When we estimate the SEM, we divide the SD by √(N-1) because sample SD is biased & consistently overestimates/underestimates the population SD
underestimates
Null hypothesis for independent samples t-test
μ1= μ2
The mean of the distribution of differences becomes zero only when all pairs of samples represent a single population.
True
The standard deviation can only be computed with reference to the mean
True
You can ONLY use a z test when you know both the population mean and population standard deviation
True
In a paired samples t-test, -2r(SEM1)(SEM2) in paired samples t-test causes the t-ratio to be (smaller/larger) than it would if the term were excluded.
larger
The further the difference of fe from the fo, the (smaller/larger) chi square, making it (more/less) likely to reject the null hypothesis
larger, more
140. Assuming that the mean differences & SDs are constant—you are more/less likely to reject the null when using a paired samples t-test than an independent samples t-test.
less
For a given sample size, setting the alpha error at .01 rather than .05 makes it __________ (more or less) likely to reject the null hypothesis
less
The smaller the t for chi square, the (more/less) likely to reject the null hypothesis
less
With extremely small samples, does the opportunity for rejecting H0 become more or less likely
less likely
The more cells, the (more/less) probability of variance by chance
more
The central limit theorem states that the shape of the entire distribution of sample means will tend to approximate
normality
The two samples represent a single population
null hypothesis
A point estimate is a hypothesized parameter value that is based on a(n) __________ value
sample
The standard error of difference is literally the standard deviation of which important sampling distribution?
sampling distribution of the difference
The t ratio tells us by how much the difference between the sample means deviates from a mean difference of zero in units of what?
standard error of the difference
If successive random samples are taken from the population, the standard deviation of the resulting distribution of means is called
standard error of the mean
Any measurement made on a sample is called a(n)
statistic
For a given value for the sample's standard deviation, increasing the sample size decreases the estimated standard error of the mean.
True
In any distribution, the median must fall at the 50th percentile
True
The larger the sample size, the value of t is almost the same as the value of z
True
The larger the standard deviation of the sample, for a given n, the larger is the estimated standard error of the mean.
True
When all pairs of samples come from a single population, the mean of the distribution of differences assumes what numerical value?
0
With a confidence interval of .95, the probability of not including the parameter mean is equal to __________ or less.
5%
The probability of selecting a sample whose mean is higher than the population mean is
50%
The probability of selecting a sample whose mean is lower than the population mean is
50%
The alternative hypothesis for the t test always specifies the direction of the difference
False
The distribution with the largest range necessarily has the largest standard deviation
False
As the degrees of freedom increase (as sample size get larger), the t distributions approach what other important distribution?
Normal
The statement that the sample is representative of the same parameter as that assumed for the population at large is called the __________ hypothesis
Null
Standard Deviation
average variation from the mean in a set of measurements in a sample
A nonrepresentative sample is often due to __________, which is constant sampling error in one direction
bias
When dependent variable is NOMINAL, we use the ___ test
chi square
The range of mean values within which the true parameter mean is predicted to fall is called the
confidence interval
Does setting alpha at .01, rather than at .05, increase or decrease the likelihood of rejecting H0?
decreases
For a single sample t-test, If sample SD increases, the obtained t-value (increases/decreases) which makes it (more/less) likely to reject the H0.
decreases, less
An estimate of the standard deviation of the entire distribution of differences
estimated standard error of the difference
Every sample measure is assumed to contain bias
false
If sample sizes are large, bias is eliminated
false
T/F The degrees of freedom would be the same for an independent samples t-test and a paired samples t-test with the same number of subjects.
false
The null hypothesis for the z test states that the hypothesized population mean is different from the known population mean
false
The standard deviation can never be zero
false
When the two sample means differ, no matter by how much, we always reject the null hypothesis
false, it has to be greater than
Changing the confidence interval from .95 to .99 (increases or decreases) the range of mean values within which μ is assumed to fall.
increases
That branch of statistics wherein estimates of the characteristics of the entire group are made on the basis of having measured a smaller group is called
inferential statistics
To calculate a t ratio, the data must come from which scale of measurement?
interval/ratio
A t-score is an estimated z-score
true
If you have larger sample sizes, there is a greater likelihood of rejecting the null hypothesis (less sampling error).
true
If you reject a null hypothesis, there is probably a significant difference
true
In z-scores, the mean is always 0 and the SD is always 1
true
T/F The reason there is only one number shown for degrees of freedom in any t-test is because the dfB must always be 1
true
T/F Whichever side the mean is on, it is skewed in that direction
true
The effect size may only be calculated when the null hypothesis has been rejected
true
The mean is to central tendency as the standard deviation is to variability
true
The standard deviation can never be greater than the range
true
To calculate the z test, the population's parameter values must be known
true
every sample measure is assumed to contain sampling error
true
Characteristics of a normal distribution
•Its mean, median, and mode are all equal •Its form is mesokurtic (the SD=1/6 of the range) •The areas under the curve are fixed (e.g., 68.26% of the area lies between +1 and -1 SD of the mean)
Characteristics of the sampling distribution of the mean
•Mean, median, mode must ALL equal population mean •Distribution of means must be NORMAL even if actual population distribution from which samples are taken is not •SD=SEM