Research methods Flashcards
What is the mean
average, calculated by adding up all the numbers and dividing by the total of numbers.
what is the median
the middle value in an ordered list
what is the mode
most common value
What are the levels of measurement used?
nominal, ordinal, interval and ratio
What is nominal level of measurement
putting into categories, usually when they are word labelled.
It provides less information than the others.
What is ordinal level of measurement
positions or ranks within a group, not with equal distance between units.
what is interval level of measurement
interval data in equal agreed units.
It provides the most information.
What is ratio level of measurement
measuments where there is equal distance between units and a true 0 point.
what are the measures of central tendency
Mean , Median , Mode
what are the measures of dispersion
Range and Standard deviation
what is the range
difference between the highest and lowest
what is standard deviation
calculates the average distance from the mean of all scores. it considers and the scores.
what is a correlation
a relationship between two variables or -1 to 1 with 0 being no correlation, -1 being a negative correlation and 1 a positive correlation
what is a positive correlation
as one variable rises so does the other
what is a negative correlation
as one variable rises the other falls
what letter stands for the correlational coefficient
r
what is a correlational coefficient
a number ranging from -1 to 1 showing the strength of a relationship. e.g. 0.76
lab experiment
a set up experiment in a labratory.
>controlled, scientific
+-tight control over variables and is replicable. high internal and external validity
- - artificial so lacks ecological validity and low mundane realism and demand charcteristics
- ethics
Iv/DV are operationalised so dosent represent reality
Field Experiment
exploitation of environment in a pubic domain.
> natural environment outside Lab
+- High Ecological Validity, high mundane realism and low demand characteristics
– low control on variables and ethical? deception
-low internal validity
Natural Experiment
capitalising a naturally occuring event which can never be replicated
+ ethical as would be unethical to set up anyway
+ high eco val. and high mundane realism
+ theory development
- cannot generalise
- rare events
- no control
Quantitative Data
numbers etc.
+- easy , neat, ecomomic, objective, can be translated into graphs etc
Qualitative Data
word etc.
–subjective, open to interpretation, partial, bias, too much data, uneconomic.
What is content analysis and how do you do it
turning qualitative data into quantitative.
1) sorting by theme. e.g. count or measure
2) coding system = frequency data
what is wrong with self report techniques
> self desirability bias
>demand characteristics
what is an closed question
a questions where there is a choice of answers
what is an open question
a question where detail and elaboration is needed.
Evaluate correlational research
ADV: correlational analysis shows the direction and strength of relationships so the findings can be used to generate ideas for future research
DISADV: correlational research only explains the two variables that are being studied, but other factors may be involved that were not known of, or accounted for.
natural observation
> observe not interfere \+ high eco. val. \+ no demand characteristics \+ theory development - extraneous variables - observer bias
content analysis
converting Qualitative to quantitative
> coding units used
> statistical analysis the carried out
what are pilot studies
small scale study before actual to foresee problems
how to control variables
- Counterbalancing- mixing up order of tasks to solve order effect
- Random allocation- to create equal balance in the group
- Extraneous variables- all variable are kept constant for all participants; same place, same time, same conditions
- Standardizing instructions- keep everything as similar as possible from context to instructions.
independent groups design
• Different participants (one with audience and one without) in each group
+ Avoids order effect, learning effect and fatigue effect
- Participants effect; one group may be better
- Twice as many needed
Repeated measures design
• All participants do both tasks (on with audience, one without)
+ Participants variables reduced
+ Fewer participants needed
- Order effect
Matched pairs design
• Different participants in each condition, but they are matched on important variables. \+ No order effect \+ Participants variables minimized - A lot of participants - Time- consuming
selecting and using pps
> target group is part of population you are interested in
> target group should be representative- if not it is biased
random sampling
every member has equal chance of being selected.
+ Fair
- Not representative- subgroups may not be selected.
opportunity sampling
participants are picked if they are available and willing.
+ Quick, easy and practical
- Not representative- can’t generalize.
volunteer sampling
participants actively volunteer.
+ Willing participants ==> more in-depth analysis.
- Participants are more likely to be more cooperative ==> not representative.
participant effect
- If a participant knows they are being observed; they will change their behavior.
- This will effect validity or reliability
social desirability bias
- participants will present themselves in a better way
demand characteristics
when participants form an idea about the study and try to ‘please’ the researcher.
researcher bias
• The researchers expectations influence their design and conduction of the study. Their behavior may produce demand characteristics
interviewer effect
may lead to leading questions, participant’s answers may be adapted to fir expectations, and interviewers behavior can affect the participants.
directional hypothesis
stating a directional prediction e.g. group A will do better than B
non- directional hypothesis
states there will be a difference, but does not state what the difference will be e.g. one group will do better than the other
null hypothesis
what you assume is true before the study; usually there will be no relationship
experimental hypothesis
if you reject the null hypothesis you must form one; usually there is a relationship
evaluate questionaires
\+ Lots of info \+ Quick, easy, cheap \+ Low researcher bias \+ Median and mode can be assessed \+ Allows for degrees of opinions \+ Anonymity - Bad questions = bad answers - Leading questions - Opened or closed questions provide different data and responses - Biased samples - Self- report ==> social desirability bias - Ethics- confidentiality
structured interviews
comprised of pre-determined questions, little training and easy to analyze.
Behavior is observed as it occurs; high mundane reality, low demand characteristics
+ Rich data
- Self report ==> social desirability bias
- Impractical _. Time consuming
- Ethics- confidentiality
unstructured interviews
set of discussion topics, less constrained about the conversation. New questions emerge as you proceed. \+ Rich data \+ Pilot study can be used - Cant generalize - Interviewer bias ==> to participants and study - Recollection unreliable - Confidentiality - Time consuming
case studies
• Intense description on a single case \+ Rich data \+ Unique ==> promote new studies or theories - Cant generalize - Ethics - Causal relationship cannot be assumed
ethical issues
CONFIDENTIALITY PRIVACY PSYCHOLOGICAL HARM INFORMED CONSENT RIGHT TO WITHDRAW DEBRIEF DECEPTION ANIMAL RIGHTS
participants observations
researcher participates in activity
+ Develops relationship
- Lose objectivity which lead to social desirability bias
non- participant observation
researcher observes
+ Greater objectivity
- Low sense of group dynamic
structured observation
pre-defined behavior categories used
Categories must be operationalised and must be objective
+ Relevant data
- Interesting data unrecorded
participant effects
Participants respond to the info they receive and try harder to give desired answer. Opposite can happen if participants aren’t interested in study
