Test #3
Conceptual Replication
A replication study in which researchers examine the same research question (the same conceptual variables) but use different procedures for operationalizing the variables
Quota Sampling
Identify sub groups of interest, establish how many of each sub group desired (quota), then collect that amount for each -ensures that subgroups of interest are included
What happens without sub groups?
No data in important sub groups
Practical Significance
The usefulness or everyday impact of the results
Non-Probability Sampling
-Convenience Sampling -Quota Sampling
Probability Theory
-Involves a study of phenomena characterized by uncertainty -Used to determine how likely outcome are to occur given what we know and don't know (how likely to win the lottery) -determine if studys sample population represents population: 1. Use good sampling techniques 2. Check samples distribution of scores to see if they are normally distributed -we can never be certain - research never 'proves' -> it supports
Probability Sampling
-Simple Random -Stratified Random
Hypothesis Test
-determines the likelihood that the results found (with this sample) were due to simple chance or error -If the likelihood is very low, then the results are considered "real" in that there is an actual effect, relationship, or difference in the population
Null Hypothesis (H0)
-hypothesis in which there is no effect -Any effect, change, or relationship found is just a chance of error
Alternate Hypothesis (Ha)
-typically our hypothesis that says there is an effect, relationship, or difference -Any effect, relationship, or change is real (not chance or error)
Factors that increase power are:
1)Larger Sample Size 2)Less error in research design 3)Larger effect size
What are the 6 steps of hypothesis testing?
1. Define Null and Alternative Hypothesis 2. Choose alpha level (almost always a=0.05) 3. Determine Critical Value(s) for rejecting null 4. Calculate Test Statistic 5. Compare Test Statistic with Critical Values 6. Accept or Reject Null, Interpret results
Calculating a Independent Samples T-Test in formulas:
1. Find SD 2. Find SD²pooled 3. Find SDx-x 4. t = .......
Finding Pearson's r in formulas:
1. Find the Means of each variable 2. Find the SDs of each variable 3. Find the Z-score for all scores 4.Mutiply each pair of X & Y scores for each participant 5. Add up all mutiplied pairs 6. Divide by the number of pairs of scores (N-1) = your r value ***Plotting data can give a good indication of scores
Representative Sample
A sample that closely resembles the population in important ways
Biased Sample
A sample that does not resemble the population in important ways
Sample
A subset of the population that is meant to represent the population (individuals actually involved in our study)
Power
Ability to detect an effect when it does exist in the population
Exact Replication
An attempt to replicate precisely the procedures of a study to see whether the same results are obtained
Perfect Representative Sample
An ideal to strive for that is generally never obtained
What can we do as skeptical scientists?
As Skeptical Scientists, we always start by assuming that our hypothesis is not correct
Convenience Sampling
Choosing members of a population that are convenient (i.e. easy, cheap, quick) to obtain
Probability Distribution
Curve of scores where the percent is the probability of obtaining a score within the range
Type 2 Error
Evidence for a statistically significant result is not found in our sample but there really is an effect on the population. In other words we accept the null when it is false. **increase power of study to decrease type 2 error
Type 1 Error
Evidence for a statistically significant result is found in our sample but there really is no effect on the population. In other words when the null is true but we reject it. ** more stringent p-value p < .001 decreases type 1 error
Non Response Bias
Extent to which those who were selected and did participate differ from those who were selected and did not participate
One Sample T-Test or One Sample Z-Test
Inferential Statistic that compares a sample mean to a known population mean to determine if they are different
Null until suggests Alternative equals ...........
Innocent until proven guilty
When something falls in the region of acceptance.....
It is NOT statically significant, reject the alternative and accept the null.
When something falls in the region of rejection.....
It is statically significant, reject the null and accept the alternative.
Statistical Significance
It is unlikely that the results were due to chance or error such that we instead consider there to be a real effect, relationship, or difference in the population. *****Statistical significance does not necessarily imply that the results are important or practically meaningful
Stratified Random Sampling
Random sampling of each groups and combining this into a bigger sample
Descriptive v.s. Inferential Terms
Sample. Population Statistic. Parameter Mean(M). Mu (μ) Standard Deviation(SD). Sigma (σ)
Pearson's r (correlation coefficient)
Statistic used to describe a linear relationship between variables
Independent Samples T-Test
Statistic used to test whether there is a statistically significant difference between 2 independent groups on their mean value of a variable
Effect Size
Strength/Size of the effect, relationship, or difference
Convenience Sampling Strengths and Weaknesses
Strengths: Convenient Weakness: (Most) Likely a biased sample
Quota Sampling Strengths and Weaknesses
Strengths: Easy, cheap, quick, and may be more representative than convenience sampling Weakness: Most likely a bias sample (to a degree)
Sampling Error
The difference(s) between a sample and it's corresponding population (Generally the larger the sample, the more accurately it represents the population)
Population
The entire group we are interested in understanding
Sampling
The process of selecting individuals to participate in a study
What does p > .05 mean
This is the "alpha level". It is the level of statistical significance. The probability of getting the results by chance is less than 5 in 100.
What are the two types of errors in hypothesis testing?
Type I & Type 2
When do you use a one sample T-TEST?
Used when N < 30 and σ is UNKNOWN
When do you use a one sample Z-TEST?
Used when N > 30 and σ is KNOWN
Critical Value (crit) (Zcrit)
Value of a statistic that defines the region for rejecting the null ***In psych, is often plus or minus 1.96 SD in a normal distribution for a z-tail test
Non-Probability Sampling
We don't know the probability of each member of the population being chosen -(mostly) biased sample
Probability Sampling
We know the probability of each member of the population being chosen -Representative sample
Simple Random Sampling
each member of the population has an equal chance of being selected Strengths: Selection is fair and unbiased Weakness: Can be difficult, expensive, or just plain impossible to do
What is the alpha level (α)?
maximum probability/risk allowed for concluding that an effect, relationship, or difference exists when in fact, it does not
What is often the critical value for a normal distribution z-tail test?
±1.96 SD
When conducting a one sample z-test the crit values are always:
±1.96 SD
In psych what is the alpha level most commonly?
α=.05