⭐ • Research Methods: Correlation, Measurement, Design & Hypothesises Flashcards
What is correlation?
A method to assess to which degree 2 variables are related or have a relationship
Correlation does not = C______________
Correlation does not = Causation
What type of data do you need to be able to formulate a correlation?
Quantitative
In correlation, you are looking for a ____________ between ____ variables
In correlation, you are looking for a relationship between two variables
List two common processes that allow for correlational analysis
- Self-report: Questionnaires or Interviews
(Questionnaires are better as in interviews, social desriability = skew)
Correlational analysis can only be done on ____________________ data as it is a ________________ process
Correlational analysis can only be done on quantitative data as it is a statistical process
How is correlation shown graphically?
On a scattergram/ scatter diagram
What are the three categories that a correlation can be categorised into?
- Positive correlation
- Negative correlation
- Zero correlation
List 3 positives to correlation
- Determines if a relationship is significant
- Useful method to conduct** preliminary analysis**
- If it determines a strong correlation, a casual relationship can be ruled out and further research can condifenty be carried out
List 4 negatives to correlation
- It cannot show a cause or effect relationship
- Variables correlated may cause change in another e.g. watching violent video could increase aggressivenes OR more violent people are choosing to watch violent videos - we cant tell
- Methods used to measure/ collect variables may lack reliability or validity
- Interveining variables or factors can impact correlative relationships, lowering the validity of the claim on a connection
What is preliminary analysis?
The initial examination of data to gain a basic understanding before going into detial
Within correlation, what are the two variables of a relationship that are measured?
- Strength
- Direction
What is the format/ symbol for displaying a correlation co-efficient?
- r = correlation co-efficient
- Format = (r=____)
What is the correlation co-efficient?
A number between -1 and 1 that tells you the strength and direction of a relationship between variable
Explain Positive Correlation
- THERE IS A RELATIONSHIP
- As one variable increases, so does the other variable
- ( A perfect) Correlation co-efficient = +1
Explain Negative Correlation
- THERE IS A RELATIONSHIP
- As one variable increases the other variable decreases
- ( A perfect) Correlation co-efficient = -1
Explain Zero Correlation
THERE IS NO CLEAR RELATIONSHIP
* Absolutely no indication of a pattern between variables
* Correlation co-efficient = 0
What is a perfect correlation (+ give an example)
- 100% of the time, the variables in question move together by the exact same strength and directio
- e.g. a perfect negative correlation has a co-efficient of -1, meaning that one variable increases at exactly the same rate as the other decreases
Which of the following shows a negative correlation?
Which of the following shows a positive correlation?
Which of the following shows a zero correlation?
What word should you never use when describing the results of a correlation?
Prove
* Never use this word when describing correlation, correlations do not allow researchers to establish a cause and effect, therefore there is never any ‘proof’
What is the design?
The specific approach used to scientifically collect, analyze and then interpret dat
What are the 3 types of design an experiment can have?
- Independent Groups
- Matched Pairs
- Repeated Measures
What is an Independent Groups design?
When you have 2+ groups with different participants and you compare them together e.g 12A compared to 12B
What are the positivies and negatives to an Independent Groups deisgn?
- +VE = Practically, least time consuming
- +VE = Less likely to get practice or fatigue effects as participants not repeating anything
- -VE = No control over individual/ environmental differences
What is a Matched Pairs design?
When you have 2+ groups with different participants and you compare them together BUT both groups match people on individual differences e.g. age, sex, etc… e.g. Raine matching his 2 groups by having same ammount of schizophrenic participants
- Attempting to negate -VE of Independent Groups by pairing people very closely on individual differences
What are the positivies and negatives to an Matched Pairs deisgn?
- +VE = Much more controlling, increased validity due to control over individual differences
- -VE = Very time consuming due to requirement to exactly match participants
What is a Repeated Measures design
When you have only 1 group, but each participant provides multiple (2+) data points. A single group but repeating measurements for same individuals
What are the positivies and negatives to an Repeated Measures deisgn?
- +VE = Controlled completely for individual differences as only one group
- -VE = Participants suceptable to practice or fatigue effects (As due to same people repeadedly doing tests, results could become worse as participants would become more tired OR participants could produce better results as they have dont it so many times before) - Lowering Validity
What are the three levels of data?
- Nominal
- Ordinal
- Interval
What is Nominal Data?
The most basic level of measurement. It’s used when data is put into tally charts/ categories. For this reason, its sometimes referred to as category data
What are the positivies and negatives to Nominal Data?
- +VE = Very easy to analyse due to it being so basic, usually reliable
- -VE = Gives very little information as it is basically a headc count, it only tells us how many people are in each group
What is Ordinal Data?
This is used when data can be put into order e.g. 1st, 2nd, 3rd (ranking data). Although, it cannot tell us the distance between rankings/ units of measurements are not definable sizes e.g. between 1st and 5th, therefore can’t tell us any relation between the numbers
(‘Ranking’ data - we used on self-reported levels of aggression)
What are the positivies and negatives to Ordinal Data?
- +VE = providing information about preferences, rankings, and degrees of something (e.g. degrees of aggression)
- -VE = Usually based on opinion therefore it tends to be subjective rather than objective
What is Interval Data?
The most complex level of measurement. There is an equal gap between each unit of measurement e.g. like in cm, where the gap between 2 and 3cm is identical to that of between 10 and 11cm (Having interval scales)
What are the positivies and negatives to Interval Data?
- +VE = Not subjective, completely objective as any person can get the same, standardised result
- -VE = The lack of a true zero. In the interval scale, there is no “true” starting point. They do not start at 0 even though one value is labeled zero. This means that zero does not signify an absence in our dataset
What is a true zero?
A true zero refers to a scale where 0 indicates the absence of something. An interval scale lacks a true zero. Examples of scales without a true zero include temperature as 0° doesnt mean there is no temperature
What are interval scales?
Interval scales measure data with equal distances between points. This means that the difference between 10 and 20 is the same as the difference between 30 and 40
What is a Hypothesis?
A statement of what you belive to be true
What are the 2 types of Hypothesis?
- H1 - Alternative Hypothesis
- H0 - Null Hypothesis
What is an Alternative Hypothesis?
(+example)
A hypothesis that infers that the results gathered will yield a significant result/ relationship/ correlation
e.g. There will be a significant relationshuip between X & Y
What is an Null Hypothesis?
(+example)
A hypothesis that infers that the results gathered will not yield a significant result/ relationship/ correlation
e.g. There will not be a significant relationshuip between X & Y
When stating a Null Hypothesis, what also must you state?
That any relationship shown is purely due to chance
What should you always specifiy/ include alongside your hypothesises?
The direction of your hypothesis
What is a Directional or One-tailed Hypothesis?
- A hypothesis that states the kind of difference between the two conditions or two groups of participants e.g. higher or lower, faster or slower etc..
- It infers the expeirment will go one way
What is a Non-directional or Two-tailed Hypothesis?
- A hypothesis that states that there is as difference between two conditions or groups of participants, but it doesnt explicitly state what said difference is
- It infers the expeirment will go either one way
