Research methods
Validity and Reliability
Internal validity- extent to which we are measuring the right thing [effects of IV on DV]
External validity- extent to which the results of a study can be generalised and applied to real life
Things that can affect internal validity:
>Demand characteristics
>Confounding variables
>Social desirability bias- in questionnaire when ppts answer in ways to make them look socially
desirable
>Investigator effects- anything the researcher does which can impact on the ppts behaviour in the study
>Poorly operationalised variables- variables that are not clearly defined and operationalised-could
result in experimenter measuring the wrong [maths test for measuring IQ]
Ways to improve internal validity
>Control extraneous variables
>Improve design of the study
>Use standardised instructions and procedures [recordings, scripts]
>Double blind/single blind to avoid bias and demand characteristics
>Ask same questions in different ways in questionnaires to check for social desirability
External validity
Ecological validity- extent to which findings can be applied to other settings in the real world
Lab experiment- usually low EcV [can’t apply findings to non-lab settings due to high control]
Natural experiment- if artificial then may no be able to generalise to other settings
Field experiment- although in a natural setting can lack ecological validity
Face validity
>the extent to which a measure looks like its measuring the right thing
How do you assess it?
1. Researcher will check if ther tool seems to be measuring the right thing
2. Peer can check it
3. Does Not require expert knowledge
How to improve it?
1. Change/reword questions that are not measuring the right thing
Concurrent validity
>extent to which a measure is similar to another measure of the same thing
How do you assess it?
1. Compare the results from one measure to those of another previously validated measure on the same
topic [compare validated and established IQ test to compare the one you've designed]
2. Ppts are given both measures to complete and the results from both are compared- if they are similar
you can claim you have concurrent validity
How to improve it?
1. Check your questions- remove or reword questions
2. Check concurrent validity again
Reliability
>extent to which research is consistent
Ways of assessing:
1. Test-retest-ppts redo the same test at a spaced time interval and both sets of results are compared-
high correlation score= reliable
2. Inter-rater reliability- different observers rate same thing- compare results- high correlation score=
reliable
Quantitative and Qualitative data
, Quantitative data- data that focuses on numbers and frequencies which can be counted [eg
experiments/questionnaires]
Qualitative data- data the described meaning an experience which is expressed in words [eg case
studies/interviews]
Quantitative AO3
>Easy to analyse- apply descriptive and statistical tests; can make comparisons and draw conclusions
>Helps reliability
>Can be presented in graphical form for easier interpretation
>Oversimplify behaviour as lacks depth or explanation behind data
Qualitative AO3
>Rich and detailed information helps to fully understand behaviour
>Provides insight into behaviour that isn’t limited to numerical data
>Difficult to reliably analyse and draw conclusions
>Can be subject to interpretation
>Harder to compare between groups
Primary data- information that has been obtained first hand by the researcher
Statistics
Histograms:no gaps allow for continuous data- no separate groups vol of each bar measures frequency
line graph continuous data
Measures of central tendency-
Informs us about middle values for a set of data-they are averages that can be calculated in different ways:
1. Mean- midpoint of the combined values of a set of data [add the scores up and divide by how many
there are]
Advantages> uses all the data and is more representative
Disadvantages> can be distorted easily distorted
2. Median- middle value where scores are arranged from lowest to highest
Advantages> extreme scores do not affect it
Disadvantages> not all scores are included in the final calculation
3. Mode- the most frequent- in some sets of data there may be two modes [bio-modal] or not mode if
all the scores are different
Disadvantage> not very representative
Measures of dispersion
Based on the spread of scores and how far they vary from one another- two types:
1. Range- take lowest value away from highest
Research methods
Experimental methods
Aim- a statement of what the researcher intends to find out in a research study
Debriefing- a post-research interview designed to inform participants of the true nature of the study and to
restore them to the state they were in at the start of the study
Hypothesis- a precise and testable statement about the assumed relationship between variables
Independent variable[what you’re changing]- some event that is directly manipulated by an experimenter in
order to test its effects on another variable eg the DV [what you’re measuring]
Validity and Reliability
Internal validity- extent to which we are measuring the right thing [effects of IV on DV]
External validity- extent to which the results of a study can be generalised and applied to real life
Things that can affect internal validity:
>Demand characteristics
>Confounding variables
>Social desirability bias- in questionnaire when ppts answer in ways to make them look socially
desirable
>Investigator effects- anything the researcher does which can impact on the ppts behaviour in the study
>Poorly operationalised variables- variables that are not clearly defined and operationalised-could
result in experimenter measuring the wrong [maths test for measuring IQ]
Ways to improve internal validity
>Control extraneous variables
>Improve design of the study
>Use standardised instructions and procedures [recordings, scripts]
>Double blind/single blind to avoid bias and demand characteristics
>Ask same questions in different ways in questionnaires to check for social desirability
External validity
Ecological validity- extent to which findings can be applied to other settings in the real world
Lab experiment- usually low EcV [can’t apply findings to non-lab settings due to high control]
Natural experiment- if artificial then may no be able to generalise to other settings
Field experiment- although in a natural setting can lack ecological validity
Face validity
>the extent to which a measure looks like its measuring the right thing
How do you assess it?
1. Researcher will check if ther tool seems to be measuring the right thing
2. Peer can check it
3. Does Not require expert knowledge
How to improve it?
1. Change/reword questions that are not measuring the right thing
Concurrent validity
>extent to which a measure is similar to another measure of the same thing
How do you assess it?
1. Compare the results from one measure to those of another previously validated measure on the same
topic [compare validated and established IQ test to compare the one you've designed]
2. Ppts are given both measures to complete and the results from both are compared- if they are similar
you can claim you have concurrent validity
How to improve it?
1. Check your questions- remove or reword questions
2. Check concurrent validity again
Reliability
>extent to which research is consistent
Ways of assessing:
1. Test-retest-ppts redo the same test at a spaced time interval and both sets of results are compared-
high correlation score= reliable
2. Inter-rater reliability- different observers rate same thing- compare results- high correlation score=
reliable
Quantitative and Qualitative data
, Quantitative data- data that focuses on numbers and frequencies which can be counted [eg
experiments/questionnaires]
Qualitative data- data the described meaning an experience which is expressed in words [eg case
studies/interviews]
Quantitative AO3
>Easy to analyse- apply descriptive and statistical tests; can make comparisons and draw conclusions
>Helps reliability
>Can be presented in graphical form for easier interpretation
>Oversimplify behaviour as lacks depth or explanation behind data
Qualitative AO3
>Rich and detailed information helps to fully understand behaviour
>Provides insight into behaviour that isn’t limited to numerical data
>Difficult to reliably analyse and draw conclusions
>Can be subject to interpretation
>Harder to compare between groups
Primary data- information that has been obtained first hand by the researcher
Statistics
Histograms:no gaps allow for continuous data- no separate groups vol of each bar measures frequency
line graph continuous data
Measures of central tendency-
Informs us about middle values for a set of data-they are averages that can be calculated in different ways:
1. Mean- midpoint of the combined values of a set of data [add the scores up and divide by how many
there are]
Advantages> uses all the data and is more representative
Disadvantages> can be distorted easily distorted
2. Median- middle value where scores are arranged from lowest to highest
Advantages> extreme scores do not affect it
Disadvantages> not all scores are included in the final calculation
3. Mode- the most frequent- in some sets of data there may be two modes [bio-modal] or not mode if
all the scores are different
Disadvantage> not very representative
Measures of dispersion
Based on the spread of scores and how far they vary from one another- two types:
1. Range- take lowest value away from highest
Research methods
Experimental methods
Aim- a statement of what the researcher intends to find out in a research study
Debriefing- a post-research interview designed to inform participants of the true nature of the study and to
restore them to the state they were in at the start of the study
Hypothesis- a precise and testable statement about the assumed relationship between variables
Independent variable[what you’re changing]- some event that is directly manipulated by an experimenter in
order to test its effects on another variable eg the DV [what you’re measuring]
