REVISION Flashcards
What is research?
A systematic way of asking questions, a systematic method of enquiry
What are the characteristics of research?
systematic logical empirical reductive replicable
Name the 4 types of research
exploratory
descriptive
explanatory
predictive
what is exploratory research?
Used when there is little or no prior knowledge of a topic.
To become familiar with basic facts, settings, and concerns; To create a general picture of conditions; To look for clues about a phenomena.
Generate new ideas, concepts, or hypotheses.
Formulate and focus questions for future research.
what is descriptive research?
Provide a detailed, highly accurate picture
Locate new data that contradict past data; To categorise or classify or clarify a sequence of steps or stages; Report on the background or context of a situation
Little attempt to explain the results
what is explanatory research?
Is involved in explaining why something happens; Looking for causal relationships between concepts; Link issues with a general principle
Extend a theory
Support or refute an explanation or prediction
Determine which of several explanations is best
what is predictive research?
forecast future phenomena, based on findings suggested by explanatory research
who carries out research?
academics, government agencies, commercial companies, consultants, managers
name the steps of the traditional method (inductionism)
i) Observation and Experiment
ii) Inductive generalisation
iii) Hypothesis
iv) Attempted verification of hypothesis
v) Proof or disproof
vi) Knowledge
what is the problem of observation in inductionism?
physical characteristics, experience, sociological factors, culture
what is the problem of the inductive process in inductionism?
the arriving at scientific laws by repeated observation - generalising from observation is not logical
what is falsification?
where theories must be exposed to refutation (not agreement)
what are the steps of Popper’s method of falsification?
i) Problem (usually to rebuff current theory)
ii) Proposed solution (new theory)
iii) Deduction of testable propositions
iv) Tests (to falsify new theory)
v) Choose between old and new theories
what are the problems with poppers method?
Newtonian/Einsteinian physics?
Limitations of observations (objective/subjective?)
Human subjective inferences (Collins and Pinch, 1993)
Theories usually complex
Many current theories would have been discarded
what was Thomas Kuhn’s revolutions? (1962)
he suggested that science can be split into normal science and revolutionary science
what is normal science?
Increases precision of measurement
Adds additional clauses to already known theories
Is ‘conservative’
what is revolutionary science?
When anomalies to current theories become too ‘serious’
‘Crisis’ and new paradigm/concepts
Science moves forwards in leaps? (e.g., theory of evolution)
what is Imre Lakatos’ research programmes?
More ‘realistic’ view of Science
‘Naive’ induction/falsificationism too simple
Scientists do not abandon theories because of one or two
falsifications
Scientific progress is a contest between rival theories (and
modifications) with the world as a referee
Long time period to decide
what are the steps of the scientific method?
i) Define Problem
ii) Read previous literature
iii) Define Hypotheses (Null hypothesis “no difference” “no effect”)
iv) Test the theory (try to falsify null hypothesis)
v) Discuss results in comparison to competing theories
vi) Conclude which theory is best
what is the aim of experimental design?
aim to control confounding variables
what are confounding variables?
- can cause false results
- alter systematically with IV so experimenter not sure whether changes in DV are due to IV or confounding
- controlled by experimenter
- E.g researching whether a lack of exercise has an effect on weight gain, confounding variable would be food consumption
what are the characteristics of a lab experiment?
well controlled but lacks realism
what are the characteristics of a field experiment?
realistic but less well controlled
Describe Independent Groups Design - Randomised Controlled Trial (RCT)
Subjects divided into groups by strict randomisation; Check that the two groups do not differ significantly (by chance) in age, weight etc. Potentially confounding variables.
describe Independent Groups Design - Pre-existing groups
E.g. elite v novice, patient v control, subjects selected who are on average similar on confounding variables.
describe Matched-Pairs Design – Pre-existing groups design
E.g. Each participant in a group of adolescents (10-20 yrs) with hypermobility is matched to an adolescent of the same gender, age and social-economic background.
E.g. Physical activity levels in adolescents with and without hypermobility (survey).
describe repeated Measures Design
Same subjects are tested under each condition. Subject is own control!
describe design with combination of multiple groups and multiple conditions
E.g. Novice-expert differences in the effect of anxiety on penalty accuracy.
This can be described as a 2 (pre-existing groups) x 2 (conditions) design.
how d you control for fatigue, learning effect, change in concentration etc in repeated measures?
- randomisation of conditions
- Quasi randomisation
- Counterbalancing
Give an example of quasi randomisation
e.g., first participant through the door does A first, next participant does B first
name the 2 specific sources of error by participants knowing which group they are in
Hawthorne effect
John Henry effect
what is the hawthorn effect?
experimental group perform better as they know they are being tested
case of a confounding variable
due to the personal interaction between experimenter and experimental group subjects
what is the John Henry Effect?
control group tries to be better and succeeds
what is a blind study
Experimenters know which participants receive the manipulation and which the placebo; the participants do not.
what is a double blind study
Niether experimenters nor subjects know who is receiving the manipulation / placebo
what is a placebo used for
To control for socio-psychological subject-investigator effects, the control group often gets a Placebo.
The placebo has all the characteristics of the manipulation, apart from the one aspect that is investigated.
name sources of error (experiment)
selection bias assignment bias rating or halo effect subject researcher interaction experimenter bias instrumentation statistical regression post hoc error
what is validity
a test is said to be valid if it measures what it claims to measure
define homoscedasticity
error is constant
define heteroscedasticity
error increases/decreases with size of variable measured
what is internal validity
did the manipulation of the IV really cause the change in the DV
Give some threats to internal validity
selection - is the control group comparable to experimental
outside events- did something happen between measuring in one condition and measuring the other
repeated testing - learning effect
loss of concentration
instrumentation
drop out
experimenter vias
what is external validity
can you generalise your findings to other people, situations tc
what is ecological validity
a time of external validity
do lab based results generalise to the real world
what is content related validity
appropriate content
what is criterion related validity
relationship to other measures
what is face validity
does the test appear to test what it aims to test
what is concurrent validity
does the relate to an existing similar measure
what is construct validity
does the test relate to underlying theoretical concepts
what is predictive validity
does the test predict later performance on a related criterion
what is reliability
a measure of whether a test would give the same results if it was administered to the same sample at 2 different times
how do you establish reliability
stability - tet retest separated by time interval
alternate forms - 2 equivalent forms of the test given to same subjects
internal consistency - test split in to 2
what is the Bland Altman plot
a method to compare 2 measurement techniques. the differences between the 2 techniques are plotted against the averages of the 2 techniques
looks at the mean of the 2 numbers and plots that against the difference between the 2 number
middle red line - systematic bias
outer red lines- mean
what is inter rater reliability / objectivity
a measure f how objective the results are
if 2 experimenters carried out the same test on the same sample, would the same results be obtained
what are descriptive statistics
measures of central tendency
measures of data spread
distribution of the data
what are the measures of central tendency
mode
median mean
what are the advantages and disadvantages of the mode
Advantages
Simplicity; a real score
It can be used with categorical variables (e.g. female/male)
Disadvantages
It may lie completely at one end of the data (not in middle)
It does not allow you to do any further calculations
There may be more than one mode (bimodal/multimodal
what are the advantages and disadvantages of the median
advantages
simplicity
extreme values do not affect it
if the distribution scores is skewed, median is more useful that mean
disadvantages
relatively strongly affected by scores in the middle
what are the advantages and disadvantages of the mean
advantages
insiders all scores
good for interval. ratio data
disadvantaes
nobody may get the mean score
mean score may not exist
sense to outliers
what are the m measures of data spread?
range, interquartile range, variance, standard deviation, coefficient of variation
how do you work out the interquartile range
distance between Q1 and Q3 representing the boundaries in the middle 50% of the distribution
what are the advantages and disadvantages of the interquartile range
ineffective by outliers
not influenced by every score in a distribution
how do you calculate variance
find the mean
take away the mean from each number
square all these numbers
add them up divide by n - 1 (this is the case in most research)
what is the coefficient of variation
describes data measured on the interval or ratio scale
standard deviation / mean x 100
it is a normalised measure of spread
what is negative skew
scores clustered to the high values
e.g an easy test
what is positive skew
scores clustered to the low values
what is kurtosis
the peakedness of the distribution
what is positive kurtosis
distribution peaked with long thin tails
what is negative kurtosis
distribution flat with lots of extreme cases
how do you calculate standard deviation?
i) Find the mean
ii) Subtract the mean from each score
iii) Square each difference score found in (ii)
iv) Add all the squares
v) Divide the total by (number of scores - 1)
vi) Square root the result
(THE SQUARE ROOT OF VARIANCE)
what is random sampling?
each individual in the target population is equally likely to be included in the study
what is stratified sampling?
population is divided then randomly sampled
what is systematic sampling?
predetermined protocol based on a random starting point and fixed ration
e.g every 10th person through the door
what is quota sampling
like stratified, but researcher selects and keeps asking people until quota is achieved.
what is snowball sampling?
existing participants
when would you carry out a power analysis
to calculate the required sample size
power = chance of finding a significant effect given that it exists in the population
what is a type 1 error
rejecting a true null hypothesis
what is nominal data
numerals represent cateogories
what is ordinal data
numbers indicate rank order of obersavtions
e.g likert
what is interval data
order and equal intervals between numbers nut not related to true 0
what is ratio data
numbers represent units with equal intervals measured from true 0
what is a hypothesis
an expected result which is testable
what are statistical tests?
tests the hypothesis
the samples are drawn from different distributions in the target population
statistical tests looks at difference in means and spread of the sample distribution
what is a false positive (type 1)
there is no real difference but you conclude there is a real difference
you make an error in rejecting the null hypothesis
when the p value is smaller than alpha..
then you reject the null hypothesis
there is a real difference
what are false negative results
type 2
you accept the null hypothesis but you should have really rejected it and accepted the experimental hypothesis
what does H0 and H1 refer to
H1- samples are coming from the same population
H1 - samples are coming from a different population
what are the advantages and disadvantages of parametric tests?
more powerful, sometimes easier to understand
more sensitive to outliers
what are parametric tests used for?
to test for differences
data must have been measured on parametric scale (interval ratio)
data must have equal variance
sample must have been randomly chosen
sample size must be almost equal
what is the t test
most common parametric test
compares the differences between means of 2 samples
How do you check if data has equal variances?
Levenes test
What was the Numemberg code for
post WW2 and war crimes trials the nuremberg code was created
principles ensure subjects are not abused/mistreated or misled
List the points in the Nuremberg code
The voluntary consent of the human subject is absolutely essential
The experiment should be such as to yield fruitful results for the good of society, unprocurable by other methods or means of study, and not random and unnecessary in nature.
The experiment should be designed and based on the results of animal experimentation and a knowledge of natural history of any disease or other problem under study that the anticipated results will justify the performance of the experiment.
The experiment should be conducted in such a way to avoid all unnecessary physical and mental suffering
No experiment should ever be conducted where there is an a priori reason to believe that death or disabling injury will occur
The degree of risk taken should never exceed that determined by the humanitarian importance of the problem to be solved by the experiment.
Proper preparations should be made and adequate facilities provided to protect the experimental subject against even the remote possibilities of injury, disabilities or death.
The experiment should only be conducted by qualified persons. The highest degree of skill and care should be required through all stages of the experiment
.
During the course of the experiment, the human subject should be at liberty to bring the experiment to an end if he/she has reached the physical or mental state where continuation of the experiment seems to him/her impossible.
During the course of the experiment the scientist in charge must be prepared to terminate the experiment at any stage, if he/she has probable cause to believe, in the exercise of good faith, superior skill and careful judgement required of him/her, that a continuation of the experiment is likely to result in injury, disability or death to the experimental subject.
what is the declaration of Helsinki (1964)
created by the world medical association for human research
provides additional info relating to informed consent, novel protocols etc
what does a research ethics committee include
Senior executive of Institution
Representatives for each of faculties/colleges
Representative of consultant staff of local hospital
Health and Safety representative
Member of the lay community
Member of legal community
Statistician
what factors does a REC consider
participants recruitment strategies informed consent scientific background procedure protection of privacy
what forms can medical screen take place
Hospital screening
GP letter
Questionnaire (e.g. PAR-Q)
Interview
why does an experimenter carry out a risk assessment
as part of the ‘duty of care’
consider any likely to unlikely risks to the health of the participant and or the experimenters
when is parental permission needed
if the subject is younger than 16
define fabrication
reporting of non existent data
define falsification
selective reporting of data
define plagiarism
duplicate publication
define correlation
statistical testing of relationship
what are correlation coefficients used for
to quantitatively describe the strength and direction of the relationship between variables
good indicator goodness of fit of the regression line
when is pearsons and spearmans rho used
Pearson product moment coefficient of correlation - both A and B on interval/ratio scale (parametric)
Spearman rho - when data are ranked (non parametric)
what is linear regression
aims to predict dependent or criterion variable from the level in independent or predictor variable
what is regression analysis
fits a model to the data to predict values of the dependent variable from the independent
what is Bivariate regression
if only 2 variables are plotted, the model is a simple straight line of best fit
what are the advantages of a questionnaire
data collection for a large sample in a wide geographical distribution in a short time
standardised forms
anonymity
particularly useful as a research methods for examining phenomena that can be assedd through self observation
what are the disadvantages of a questionnaire
potential misunderstanding or misinterpreting questions or response choices
unknown accuracy or motivation of the respondents
what does reliability refer to in a questionnaire
whether the respondents read the questions carefully and answers consistently
what is the typical advice for question order
start with factual questions, move to attitudes
start with easier quicker questions
funnel techniques- broad to narrow
filter questions - exclude from further questions
what are epidemiological approaches to research
the study of the distribution and determinants of health related status or events in populations and the application of this study to control of health problems
concerned with frequency and distribution of diseases, specific causes and risk factors
what is morbitity
incidence of a specific disease
what is mortality
incidence of death from a disease
what is prevalence
number of existing cases of disease at a given time
what is incidence
number of new cases of a disease in a specified time period
what is a risk factor
a factor which is associated with prevalence of a disease
e.g inactivity
what is relative risk
incidence rate in exposed population / incidence rate in non exposed population
what do sports injury epidemiology studies typically address
incidence of injury in particular sports
risk of injury associated with certain sports
common sites and types of injuries
cause of the injuries
participation/economical loss due to sports injury
what are prospective longitudinal severys
same group of subjects followed over time
what are retrospective longitudinal surveys
group of people are asked to recall matters from the past
look back to determine possible causes
what is a meta analysis
a statistical method of combining the result of a series of independent previously published studies carried out for the same general purpose
type of systematic literature review
explain the procedure of a meta analysis
research question define selection criteria identify all relevant literature read full text and evaluate extract study characteristics calculate effect size apply appropriate statistical techniques to analysis the statistical significance of the results report all these steps and outcomes
what are the advantages of conducting a meta analysis
increase power by increase sample size
improve estimates of edict size
resolve uncertainty when conflicting results occurred
improve the generalisability of findings
allows reviewer to find and compare previous studies
allows reviewer to test hypotheses, discover something new about topic
what are the disadvantages of meta analysis
disagreements over effect size calculations
effect of different sample size in different studies
What is the pearson product moment coefficient of correlation used for
correlation testing when variables A and B are on interval/ratio
what is the spearmans rho used for
correlation testing when both variables are ranked
How big is important according to Hopkins
trivial = 0 small effect= 0.2 moderate effect= 0.6 large effect = 1.2 very large = 2 nearly perfect = 4
what statistical measures are best for parametric data
mean, standard deviation
what statistical measures are best for non parametric data
median and IQR
are parametric or non parametric tests more likely to make type 2 error
non parametric (also have less statistical power, more conservative, less able to test a difference which is truly there)
what is concurrent validity
f newly developed test gives same outcomes as gold standard test, then the new test has high concurrent validity
what is face validity
does the test measure what it appears to
what is content validity
do all the items on the test adequately represent what you are trying to measure
what is construct validity
used to attribute too complex to be measured easily or that has no criterion measure
when is the pearson product moment used
parametric
correlation
both A and B interval/ratio
when is the spearman rho used
non parametric
correlation
data ranked
what is the standard error of the estimate (SEE)
indicator of the average error of prediction for the regression equation
the amount of error if you use the linear regression to predict a score
the better the fit of the regression line, the less variability around it and the smaller the SEE
what is the correlation coefficient used for
to assess the goodness of fit
