RESEARCH SUPPORT - MATHS Flashcards
evaluate quantitative data
+objective
+quicker to analyse
- lacks depth - no matter how complex score system is they cannot acyurratley reflect complex human behaviour
evaluate qualitative data
+detailed
-subjcetivee interpretation
- longer to analyse
what is primary data?
collected by researcher
evaluate primary data
+research will fit the aim
-time consuming
what is secondary data?
data found by someone else - meta analysis
evaluate secondary data
+cost effective because dont have to conudct research
+time effective
- may not fully match aims
-cabt b certain on quality of data
what are the measures of central tendency
mean
mode
median
evalauate the mean
+accounts all data
-effected by extreme values
evaluate mode and meadian
not effected by extremes
doesnt consider all data
what are the measures of dispersion
range
standard deviation
evaluate range
+simple
-extreme values
evaluate SD
+not effected by extreme values
-complicated
what is sd
shows how far from the mean the majority of the data is. Shows the extent of variance and individual differeces
what does a small sd mean?
the data is consistent and has little variance/individual differences
template for interpreting mean in exam question
the mean score on average sugests….. the mean for … was which is higher/lower than the mean score for….. This means that on average …….
template for interpreting sd in exam question
the sd suggests the spread of data was more consistent for ….
this is because the sd for … was … which is lower than the sd for … which was … This means that ….. has more/less variance and individual differences.
describe a positivley skewed curve on a normal distrubtion
the hump is closer to origin
mode is High - always at the tip of the hum
median is centre
mean is Low
describe a negativley skewed normal distrubution curve
hump is on the right
mode is high - aT THE TOP OF THE HUMP
mean = centra
median = middle of them
describe bell shaped normal distrubution curve
symetrical
mode,meadian and mean are all in the cetnre (roiughly)
describe bell shaped normal distrubution curve
symetrical
mode,meadian and mean are all in the cetnre (roiughly)
what measure of design is related?
Matched pairs
Repeated measures
what measure of design is unrelated?
independent groups
what is nominal data?
pp put into categories
not everyone gets a score
most basic
what is ordinal data?
data can be ranked
each pp gets a score
subjective scale
distance between each value is not objective
what is interval/ratio data?
each pp gets a score
uses a specific established scale
what things determine what stats test should be used?
Whether the data is looking for a difference or correlation
Whether the experimental design is related or unrelated
Whether the data is nominal, ordinal or interval
how to conduct a sign test?
1) work out how many +/-/0 there are by taking away condition b from a ( A - B)
2) cancel out the 0s
3)calculate how many of each sign
4)s value is the sum of the lowest sign
what is the s value?
in a sign test its the value of the lowest sign
what is the significance level?
probability that the data isnt a fluke
what is the standard sig level?
p < 0.05 (5%)
what does 5% sig mean?
theres less than 5% chance that the results were found by chance making it more relaibla e and valid
statement of sig formula……
- the calculated symbol of … was …….the critical value( p< ….. n=….. 1/2 tailed test) is ….
- because calculated value was greater/lower than the critical value, it was decided that the results are/arent sig
- therefore we accept/reject the null
- this means that
what is a type one error?
a false positive
reject null when it is true
you didnt find anything but possibly say you did becasue sig was high
what is a type 2 error?
a false negative
say found nothing when have
falsley accept null
possibly because p level is too low
what do psychologists use p<0.05?
to avoid type one and two errors. since it is a happy medium
what ways can you analyse qualitative data?
thematic and content analysis
outline content analysis
indirect observation - turning qualitative data into quantitative
procedure of content analysis
- decide research question
- select a sample (diary entries or somet)
- Coding - decide behavioural categories (words that appear)
- work through the sample and TALLY BCS
- data analysis can be performed
outline thematic analysis
observtaion then picking out themes
procedure of thematic analysis
1) gather data from interviews, diaries or focus groups (qualitative)
2) transcribe data and become familiar with it
3) analyse without any preconceptions
4) themes will emerg, review themes
5) themes form the codiing system
6) analyse and back up with data from the sample
7) write up report
evaluate thematic analysis
+ depth and detailed reflect complex human behaviour - more hollistic
- time consuming
- too subjective and lacks scientific rigour
- analysed outside of the context that it was written/spoken so doesnt convey true intentions of sample
what is the research process?
conduct research
write up
peer review
publish report in journal
what are the features of a report
abstract
intro
hypothesis and aims
method
data + results
discussion
reference
appendix
process of peer review
1) numerous anonymous unpaid reviewers in the same field of research at hand will asses the research by criticising it
2) theyll do this independently
3) review appropriatness of methods and design
4) can accept/reject/accept with revisions
5) editor makes final decision
purpose of peer review
- to allocate research funding
- check relevance
- validiate research
evaluate peer reviews
- bias - rivals etc
- publication bias -headline grabing title that makes the like it
- bias towards the status quo - its valid but goes against internal beliefs or the norms
what are the six features of a science?
objectivity
falsifiability
paradigm
replicablity
empericism
theory construction and hypothesis testing
what does theory construction and hypothesis testing refer to?
- ## theory must be tested
what is the purpose of a stats test?
to evaluate the correctness of an empirical hypothesis by determining the likelihood of the sample results occurring by chance.
how do you conduct a sign test?
DO CONDITION A - B
add up all +
add up all -
discard 0
the s value is the lowest of the sum of negative/positive sign
what is an abstract in a report?
a short summary that includes main elements of the research - aims, hypothesis, method, results and conclusion
what is the appendix in a report?
where all used rescources go
what is the introduction of a report?
review of the general purpose of the investigation detailing relevant theories/concepts/studies that are related to the current study
what should the method section include?
- should be detailed so researchers may replicate if they want to
1. design
2. sample
3. material
4. procedure - including debrief and standardised instructions
5. ethics considered
what should the results section include?
- summarise key findings
- descriptive stats -tables and graphs etc
- inferential stats - significance, stats test, calculated and critical values
- if it was qualitative talk about themes
what does the dicussion include?
- summarise findings verbally
- discuss in context of other research mentioned in the intro
- should be mindful of limitations of their research
- the wider applications of this research - what contributions has it had for society
what is the referencing section?
full details of any sources or materials used
may include journals or books
explain paradigm in science
- khun argued that what makes something a scientific discipline is a shared set of assumptions and methods - a paradigm
- for some psychology has too many conflicting ideas and approaches, so does not have a paradigm and cannot be considered a science
- paradigms shifts mark progress within a science e.g. copernican revolution, planets go around the sun
- could be argued psychology had a paradigm shift - neuroscience
explain the falsifiability aspect of a science?
- karl popper
- a response to verificationism
- an established science must be able to be proved false. It must have claims that can be tested and proven either true or false through empiricism.
-theories that survive attempts of falsification become the strongest - this is why alt hyp always is accompanied by a null hyp
explain replicability
- for a scientific theory to be accepted it must be replicable across a number of different contexts
- replication also determines validity
- and reliability
- and the extent to which findings can be generalised
explain objectivity and empiricism
- objective - no personal opinion or subjectivity
- must use experiences and observation
- john locke
- theory cannot claim to be scientific unless it has been empiriclally tested and verified
what are differential stats
measures of contral tendency and measures of dispersion
what are inferential stats
statistical tests
