Unit 3 Flashcards
Pilot study
Small scale preliminary study integral to development of investigation
What does pilot study evaluate
Feasibility, cost, time, statistical variability and adverse events
What are pilot studies used for (3)
To help establish number of repeat measurements required, assess validity and ensure an appropriate range of values for the independent variable
Simple experiments
Single independent variable
Multifactorial experiment
More than one independent variable or combo of treatments
In vivo
In body
In vitro
Outside body
Observational studies
If cannot control things die to ethical concerns or impracticality, just observe. Less useful for causation but good for correlation
Confounding variables
Ones outside independent variable that may affect dependent variable
Keep constant or monitored
Blocking
If confounding variables cannot be controlled their effects are minimised by selecting control groups in which effects of confounding variables are equal
Randomised experimental design
Individuals are randomly assigned to an experimental group reliable
Randomised block design
If specific differences among groups of subjects occur eg gender then divided into homogenous groups then randomly assigned to treatment group
Controls
Parallel treatment in which factor being investigated is kept constant or at standard conditions
Negative control
Negates false positive. Addition of denatured enzyme when studying effect of enzyme on substrate
Hydrogen peroxide with water
Positive control
Included to check system can detect positive result negate false positive
Hydrogen peroxide with cátales
Placebo
Treatment designed to be harmless and have no effect. No one knows that they’ve taken
Sampling types
Stratified
Systematic
Random
Random sampling
each member has equal chance of selection
Stratified
Population is divided into categories that are then sampled proportionally
Systematic sampling
Members of a population are selected at regular intervals
Representative sample must
Share same mean as whole population, share same variation as the whole population
Accuracy depends on
Quality of apparatus and skill of scientists taking the measurement. How close the data is to the actual value
Precision depends on
Choosing the correct apparatus and using ur properly. Closeness of repeated measurements to each other
Reliability depends on
Consistency of results each time the activity is repeated
Validity depends on
How robust the experimental design is, how precise, accurate and relatable the results are
Replication
A completely separate repeat of the whole experiment with new samples etc
Tank a tank b
Repetition
Different samples measured in an experiment
One person- 3 blood samples
Or
Three people- three blood samples
How to make Natural variation not a problem
Sample size large enough so that any differences between organisms are not hidden by natural variation between organisms
Types of data presentation
Qualitative quantitative ranked
Error bars
Lines through point on graph parallel to an axis which shows variation in the data
Confidence intervals
Statistical estimate of range of values within certain percentage of the population would be found
Types of correlations
Positive negative no
Strong weak very weak
Anything that cannot be proven or disproven is
Non falsifiable and do not have a place in science
Scientific cycle
Make observations and construct a hypothesis Design an experiment to test hypothesis Gather and record data Evaluate results and draw conclusions Revise hypothesis if necessary
Results must be
Reproducible
Negative results are good because they (3)
Prevent needless duplicates of results
Inspire others to alter the experiment and gain positive
Make more realistic
Why is peer review important
Assessed quality and suitability
How to maintain integrity and honesty
Unbiased presentation and open about funding source
Avoid plagiarism
Replication must be possible
Cite and supply references
3Rs
Reduce
Replace
Refine
Ethic checks for human studies (4)
Informed consent
Right to withdraw data
Confidentiality
No harm
