7.1: Correlational studies Flashcards
Experiments look for what?
Experiments look for a difference between 2 conditions of an IV
Experiments look for a difference between 2 conditions of an IV, while correlational studies involve what?
Experiments look for a difference between 2 conditions of an IV, while correlational studies involve measuring the:
1. Strength
2. Direction of relationships
between co-variables
Experiments look for a difference between 2 conditions of an IV, while correlational studies involve measuring the strength and direction of relationships between co-variables.
Example
For example, Holland’s (1967) study of the relationship between:
- Locus of control
- Obedience
A positive correlation occurs where what?
A positive correlation occurs where one co-variable increases, as another co-variable increases
A positive correlation occurs where one co-variable increases, as another co-variable increases.
Example
For example, ice cream sales increase as the temperature increases
A negative correlation occurs where what?
A negative correlation occurs where one co-variable increases, while another co-variable decreases
A negative correlation is where one co-variable increases, while another co-variable decreases.
Example
For example, raincoat sales decrease as sunny weather increases
Co-variables
Co-variables are the variables investigated in a correlation
Co-variables are the variables investigated in a correlation.
They are not referred to as the independent and dependent variables, because what?
The variables investigated in a correlation are not referred to as the independent and dependent variables, because the study is:
- Investigating the relationship between them
- Not trying to show a cause and effect relationship
Scattergraphs are also known as what?
Scattergraphs are also known as scattergrams
Scattergraphs
Scattergraphs are a type of graph used to display the extent to which 2 variables are correlated
Advantages of correlational analysis:
1. Allows what to be made?
An advantage of correlational analysis is that it allows predictions to be made
Advantages of correlational analysis:
1. Allows predictions to be made.
Predictions can be made from correlations, like predicting what?
Predictions can be made from correlations, like predicting the number of ice creams that will be sold on hot days
Advantages of correlational analysis:
1. Allows predictions to be made.
Predictions can be made from correlations, like predicting the number of ice creams that will be sold on hot days.
2. Allows what?
An advantage of correlational analysis is that it allows quantification of relationships
Advantages of correlational analysis:
1. Allows predictions to be made.
Predictions can be made from correlations, like predicting the number of ice creams that will be sold on hot days.
2. Allows quantification of relationships.
Correlations show what?
Correlations show the strength of relationship between 2 co-variables
Advantages of correlational analysis:
1. Allows predictions to be made.
Predictions can be made from correlations, like predicting the number of ice creams that will be sold on hot days.
2. Allows quantification of relationships.
Correlations show the strength of relationship between 2 co-variables.
What does a correlation of +0.9 (90% similarity) indicate?
A correlation of +0.9 (90% similarity) indicates a high positive correlation
Advantages of correlational analysis:
1. Allows predictions to be made.
Predictions can be made from correlations, like predicting the number of ice creams that will be sold on hot days.
2. Allows quantification of relationships.
Correlations show the strength of relationship between 2 co-variables.
What does a correlation of -0.1 (10% similarity) indicate?
A correlation of -0.1 (10% similarity) indicates a weak negative correlation
Advantages of correlational analysis:
1. Allows predictions to be made.
Predictions can be made from correlations, like predicting the number of ice creams that will be sold on hot days.
2. Allows quantification of relationships.
Correlations show the strength of relationship between 2 co-variables.
A correlation of -0.1 (10% similarity) indicates a weak negative correlation.
3. No what?
An advantage of correlational analysis is no manipulation
Advantages of correlational analysis:
1. Allows predictions to be made.
Predictions can be made from correlations, like predicting the number of ice creams that will be sold on hot days.
2. Allows quantification of relationships.
Correlations show the strength of relationship between 2 co-variables.
A correlation of -0.1 (10% similarity) indicates a weak negative correlation.
3. No manipulation.
Correlations do not require what?
Correlations do not require manipulation of variables
Advantages of correlational analysis:
1. Allows predictions to be made.
Predictions can be made from correlations, like predicting the number of ice creams that will be sold on hot days.
2. Allows quantification of relationships.
Correlations show the strength of relationship between 2 co-variables.
A correlation of -0.1 (10% similarity) indicates a weak negative correlation.
3. No manipulation.
Correlations do not require manipulation of variables and so can be used where what?
Correlations:
- Do not require manipulation of variables
- So can be used where carrying out an experiment may be unethical
Weaknesses of correlational analysis:
1. What problem?
A weakness of correlational analysis is a quantification problem
Weaknesses of correlational analysis:
1. Quantification problem.
Correlations that appear low, for example +0.28, can sometimes be what?
Correlations that appear low, for example +0.28, can sometimes be significant (meaningful) if the number of scores is high
Weaknesses of correlational analysis:
1. Quantification problem.
Correlations that appear low, for example +0.28, can sometimes be significant (meaningful) if the number of scores is high, while correlations that seem high, for example +0.76, are not always what?
Correlations that:
1. Appear low, for example +0.28, can sometimes be significant (meaningful) if the number of scores is high
,while:
2. Seem high, for example +0.76, are not always statistically significant
Weaknesses of correlational analysis:
1. Quantification problem.
Correlations that appear low, for example +0.28, can sometimes be significant (meaningful) if the number of scores is high, while correlations that seem high, for example +0.76, are not always statistically significant.
2. C and e?
A weakness of correlational analysis is:
- Cause
- Effect
Weaknesses of correlational analysis:
1. Quantification problem.
Correlations that appear low, for example +0.28, can sometimes be significant (meaningful) if the number of scores is high, while correlations that seem high, for example +0.76, are not always statistically significant.
2. Cause and effect.
Correlations do not show causality, because what?
Correlations do not show causality, because they’re not done under controlled conditions
Weaknesses of correlational analysis:
1. Quantification problem.
Correlations that appear low, for example +0.28, can sometimes be significant (meaningful) if the number of scores is high, while correlations that seem high, for example +0.76, are not always statistically significant.
2. Cause and effect.
Correlations do not show causality, because they’re not done under controlled conditions.
Therefore, we cannot say what?
Therefore, we cannot say that one co-variable has caused the other
Weaknesses of correlational analysis:
1. Quantification problem.
Correlations that appear low, for example +0.28, can sometimes be significant (meaningful) if the number of scores is high, while correlations that seem high, for example +0.76, are not always statistically significant.
2. Cause and effect.
Correlations do not show causality, because they’re not done under controlled conditions.
Therefore, we cannot say that one co-variable has caused the other.
3. What relationships?
A weakness of correlational analysis is extraneous relationships
Weaknesses of correlational analysis:
1. Quantification problem.
Correlations that appear low, for example +0.28, can sometimes be significant (meaningful) if the number of scores is high, while correlations that seem high, for example +0.76, are not always statistically significant.
2. Cause and effect.
Correlations do not show causality, because they’re not done under controlled conditions.
Therefore, we cannot say that one co-variable has caused the other.
3. Extraneous relationships.
Other variables may do what?
Other variables may influence the co-variables
Weaknesses of correlational analysis:
1. Quantification problem.
Correlations that appear low, for example +0.28, can sometimes be significant (meaningful) if the number of scores is high, while correlations that seem high, for example +0.76, are not always statistically significant.
2. Cause and effect.
Correlations do not show causality, because they’re not done under controlled conditions.
Therefore, we cannot say that one co-variable has caused the other.
3. Extraneous relationships.
Other variables may influence the co-variables.
Example
For example:
1. Many holidays are taken in the summertime
2. People eat ice creams on holiday
,therefore, the variable ‘holiday’ is related to both temperature and ice cream sales
Weaknesses of correlational analysis:
1. Quantification problem.
Correlations that appear low, for example +0.28, can sometimes be significant (meaningful) if the number of scores is high, while correlations that seem high, for example +0.76, are not always statistically significant.
2. Cause and effect.
Correlations do not show causality, because they’re not done under controlled conditions.
Therefore, we cannot say that one co-variable has caused the other.
3. Extraneous relationships.
Other variables may influence the co-variables.
4. Only works for what?
A weakness of correlational analysis is that it only works for linear relationships
Weaknesses of correlational analysis:
1. Quantification problem.
Correlations that appear low, for example +0.28, can sometimes be significant (meaningful) if the number of scores is high, while correlations that seem high, for example +0.76, are not always statistically significant.
2. Cause and effect.
Correlations do not show causality, because they’re not done under controlled conditions.
Therefore, we cannot say that one co-variable has caused the other.
3. Extraneous relationships.
Other variables may influence the co-variables.
4. Only works for linear relationships.
Correlations only measure what?
Correlations only measure linear (straight-line) relationships
Weaknesses of correlational analysis:
1. Quantification problem.
Correlations that appear low, for example +0.28, can sometimes be significant (meaningful) if the number of scores is high, while correlations that seem high, for example +0.76, are not always statistically significant.
2. Cause and effect.
Correlations do not show causality, because they’re not done under controlled conditions.
Therefore, we cannot say that one co-variable has caused the other.
3. Extraneous relationships.
Other variables may influence the co-variables.
4. Only works for linear relationships.
Correlations only measure linear (straight-line) relationships.
Example
For example, correlations cannot show the relationship between:
1. Temperature
2. Aggression
,because it is curvilinear (not a straight line)
Weaknesses of correlational analysis:
1. Quantification problem.
Correlations that appear low, for example +0.28, can sometimes be significant (meaningful) if the number of scores is high, while correlations that seem high, for example +0.76, are not always statistically significant.
2. Cause and effect.
Correlations do not show causality, because they’re not done under controlled conditions.
Therefore, we cannot say that one co-variable has caused the other.
3. Extraneous relationships.
Other variables may influence the co-variables.
4. Only works for linear relationships.
Correlations only measure linear (straight-line) relationships.
For example, correlations cannot show the relationship between temperature and aggression, because it is curvilinear (not a straight line).
As temperature increases, what happens?
As temperature increases, aggression levels increase up to an optimum point
Weaknesses of correlational analysis:
1. Quantification problem.
Correlations that appear low, for example +0.28, can sometimes be significant (meaningful) if the number of scores is high, while correlations that seem high, for example +0.76, are not always statistically significant.
2. Cause and effect.
Correlations do not show causality, because they’re not done under controlled conditions.
Therefore, we cannot say that one co-variable has caused the other.
3. Extraneous relationships.
Other variables may influence the co-variables.
4. Only works for linear relationships.
Correlations only measure linear (straight-line) relationships.
For example, correlations cannot show the relationship between temperature and aggression, because it is curvilinear (not a straight line).
As temperature increases, aggression levels increase up to an optimum point.
Then any further increase in temperature leads to what?
Then any further increase in temperature leads to a decline in aggression levels
With correlations, why is there no IV or DV?
With correlations, there is no IV or DV, because:
- No experiment has been conducted
- The data has simply been compared and a relationship has become apparent
Even when a relationship is shown between 2 co-variables, why do we need to be cautious about our conclusions?
Even when a relationship is shown between 2 co-variables, we need to be cautious about our conclusions, because:
- No experiment has taken place
- So we have not controlled extraneous variables
What does the number +1 represent?
The number +1 represents a perfect positive correlation
The number +1 (what )represents a perfect positive correlation?
The number +1 (maximum strength) represents a perfect positive correlation
The number +1 (maximum strength) represents a perfect positive correlation.
What does 0 mean?
0 means no correlation
The number +1 (maximum strength) represents a perfect positive correlation.
0 means no correlation.
The nearer the number is to +1 or -1, what?
The nearer the number is to: 1. +1 Or, 2. -1 ,the stronger the correlation
The number +1 (maximum strength) represents a perfect positive correlation.
0 means no correlation.
The nearer the number (what) is to +1 or -1, the stronger the correlation?
The nearer the number (the correlation coefficient) is to: 1. +1 Or, 2. -1 ,the stronger the correlation
Because experiments are about difference and correlations are about relationships, we write hypotheses for experiments and correlations differently.
In an experiment you might compare how confident tall people are compared to short people - to see what?
In an experiment you might compare how confident tall people are compared to short people - to see if there’s a difference in confidence
Because experiments are about difference and correlations are about relationships, we write hypotheses for experiments and correlations differently.
In an experiment you might compare how confident tall people are compared to short people - to see if there’s a difference in confidence.
So the alternate hypothesis reflects this difference.
Example
For example, there will be a difference in how confident tall people are compared to short people
Because experiments are about difference and correlations are about relationships, we write hypotheses for experiments and correlations differently.
In an experiment you might compare how confident tall people are compared to short people - to see if there’s a difference in confidence.
So the alternate hypothesis reflects this difference.
For example, there will be a difference in how confident tall people are compared to short people.
Correlations look for relationships between 2 variables.
The hypothesis reflects this.
Example
For example, there is a relationship between:
- Height
- Confidence
In correlations, a non-directional hypothesis is a hypothesis that says that there will be a relationship between 2 variables, but what?
In correlations, a non-directional hypothesis is a hypothesis that:
1. Says that there will be a relationship between 2 variables
,but
2. Does not say what the direction of the relationship will be - it doesn’t say whether it will be a positive or negative correlation
In correlations, a directional hypothesis is a hypothesis that says that there will be a relationship between 2 variables and does say what the direction of the relationship will be - it does say what?
In correlations, a directional hypothesis is a hypothesis that:
- Says that there will be a relationship between 2 variables
- Does say what the direction of the relationship will be - it does say whether the correlation will be positive or negative
In correlations, what is the researcher looking at?
In correlations, the researcher is looking at relationships
In correlations, the researcher is looking at relationships, but any extraneous variables are not controlled.
What does this mean?
This means that a third unknown variable could be causing the results
In correlations, the researcher is looking at relationships, but any extraneous variables are not controlled.
This means that a third unknown variable could be causing the results.
Therefore, what?
Therefore, correlations cannot establish cause and effect relationships - they can only establish relationships
