Research Methods Experimental Flashcards
volunteer sampling strengths and weaknesses
strength = quick and easy for researcher - no researcher bias unless they choose where its advertised
weakness = not always representative ( volunteer bias ) - time consuming if no one volunteers
random sampling strengths and weaknesses
strength = not time consuming if small sample - representative if large sample - removes researcher bias
weakness = can be time consuming if large sample
systematic sampling strengths and weaknesses
strength = no researcher bias unless they pick the rule
weakness = might not be representative if small sample - can be time consuming if big sample
stratified sampling strengths and weaknesses
strength = subgroups are proportionally represented - no researcher bias
weakness = time consuming
opportunity sampling strengths and weaknesses
strength = not time consuming
weakness = not representative - researcher bias if they choose where to find the sample
lab experiment definition
highly controlled experiments where the researcher has full control over where and how the research will be conducted - researcher directly manipulates IV and precisely measures DV
field experiment definition
researcher still attempts to hold all variables constant while isolating and manipulating IV and measuring effect on DV and the research is conducted in a natural environment
natural experiment definition
conducted in p’s natural environment and IV is naturally occurring - researcher is unable to control EV’s
quasi experiment definition
the IV is within a person e.g gender, age or schizophrenia and is not manipulated by researcher - its naturally occurring but concerning a person
lab experiment strengths
-high in internal validity as researcher has high control over EV’s and directly manipulates IV which establishes cause and effect
-highly replicable as high degree of control means the experiment can be repeated and same results will be found so high in reliability
lab experiment weaknesses
-low in ecological validity as conducted in artificial environments so doesnt represent the real world
-may be influenced by demand characteristics as its in an artificial environment and are often aware they’re being studied so may change their behaviour not representing their natural actions so lacks validity
field experiment strengths
-fair degree of internal validity as researcher attempts to control EV’s and isolates IV to allow direct manipulation so change in DV is most likely a result of change in IV
-less likely to be influenced by demand characteristics as the experiment is conducted in their natural environment so p’s will act natural so high in ecological validity
field experiment weaknesses
-ethical issues as p’s are studied in their natural environment so may not be aware they are being studied so not given the opportunity to provide informed consent
-lower in internal validity compared to lab experiments as although researcher attempts to control EV’s they’re unable to do this as well as in a lab experiment so can’t be as confident in establishing causality
natural experiment strengths
-high in ecological validity as research is in a natural setting and IV is naturally occurring so p’s are less likely to change behaviour and show demand characteristics
-useful in investigating concepts that would be unethical to directly manipulate as researcher doesnt manipulate IV so doesnt break the ethical guideline of protection from harm
natural experiment weaknesses
-low in internal validity as researcher has little to no control over EV’s and the IV isn’t isolated or carefully manipulated so difficult to establish causality
-ethical issues as p’s may be unaware they are being studied so aunt given the opportunity to provide informed consent and may not be willing to participate
investigator effects meaning and how to control
results may be knowingly or unknowingly influenced by the researcher as their expectations may influence how they design a study or how they behave towards p’s which may lead to demand characteristics
ways to control = randomisation, standardised interactions and double blind technique (researcher doesn’t know what group p’s are in)
confounding variable meaning
something that varies systematically with the IV and may act as a second IV
one tailed (directional) and two tailed (non-directional) hypothesis definitions
directional = states there will be a change in the DV and what direction that relationship will go
non-directional = states there will be a change but doesnt state what direction the change will go in
random sampling definition
giving ever member of the target group an equal chance of being selected e.g random number generator
opportunity sampling definition
p’s who are freely available, accessible and willing to take part at the time
volunteer sampling definition
p’s are used that volunteered or put themselves forward to take part - involves placing an advert to inform target population
stratified sampling definition
researcher identifying sub groups of the target population and calculating the percentages that each sub group make up the target population which are then proportionally represented in the sample
systematic sampling definition
chosen for selecting from a target group e.g every 4th person in a list
independent groups definition, strength and weakness
p’s are randomly allocated to one group of the IV each with different conditions
strength = no order effects
weakness = participant variables
repeated measures definition, strength and weakness
p’s take part in both conditions of IV
strength = no participant variables
weakness = order effects
matched pairs definition, strengths and weakness
p’s are matched based on similarities or key characteristics e.g gender which are believed to have an influence on the DV - one pair in condition 1 and another in condition2
strength = no p variables + no order effects
weakness = time consuming
British Psychological Society ethical guidelines (7)
d,rtw,pfh,c,d,p,ic
deception = withholding info from p’s should be avoided
right to withdraw = p’s can withdraw themselves or their results at any time during or after the study
protection from harm = investigators have a responsibility to protect p’s from psychological or physiological harm
confidentiality = p’s data should not be disclosed to anyone unless agreed and stored securely
debrief = all relevant details of study should be given to p’s before and after study
privacy = p’s daily life is not intruded
informed consent = p’s given all details for them to make a decision whether to take part or not - parental consent for those under 16
central tendency - mean advantages and disadvantages
+sensitive as considers all data in calculation
-can be affected by anomalies and sometimes value doesnt make sense e.g 2.4 children
central tendency - median advantages and disadvantages
+is not affected by anomalies
-isnt as sensitive as doesnt include all values
-can be unrepresentative in small data sets
central tendency - mode advantages and disadvantages
+less influenced by anomalies and makes more sense then mean e.g 2 children not 2.4
-may be more than one mode in data set and less meaningful if not normal distribution
dispersion - range advantages and disadvantages
+easy to calculate so less time consuming and less prone to error
-very vulnerable to distortion by extreme values and doesn’t show data in relation to mean
dispersion - standard deviation advantages and disadvantages
+more sensitive and includes all scores rather than highest and lowest
+allows interpretation of individual scores e.g if score is in top 10%
-more complicated to calculate so more time consuming and prone to error
-less meaningful if no clear distribution
high, small and zero standard deviation meaning
large SD = a lot of variation around the mean
small SD = majority of data was clustered around mean
zero SD = all data was same
when should we use bar charts
when data is discrete (divided into categories or is non-continuous)
when should we use histograms
when data is continuous so bars must touch - scores are on x-axis
when should we use scattergrams
when looking for a correlation or a relationship between 2 variables
when should we use pie charts
used for discrete non continuous data to show percentages that contribute to the whole sample
normal distribution definiton
standard deviation (bell curve) , mean median and mode are in the middle and data is distributed evenly around it so the graph is symmetrical
negative skew distribution definition
when the mean value is the lowest value
positive skew distribution definiton
when the mean is the highest value
ordinal data definition
data that can be ranked or placed in order but the difference between values art the same
nominal data definition
data that can be placed into separate categories - usually involves counting the frequency of behaviours
interval data definition
has a standardised measurement such as time where the distance between is equal and consistent
3 factors considered when choosing a statistical test
design, data and difference or correlation
type 1 error definition
false positive as the researcher claims they have found a significant difference when there isn’t one so they wrongly reject the null hypothesis and wrongly accept the experimental - most likely to happen when significance level is too lenient
type 2 error definition
false negative as the researcher claims there has been no significant difference when there has been so they reject the experimental hypothesis and wrongly accept the null
what level of significance is normally accepted
0.05 , 5%
