3.2 Experimentation

what is validity

variables are controlled so that any measured effect is likely due to independent variable

what is reliability

consistent variables are achieved in repeats and independent replicates

what is accuracy

data is close to true value

what is precision

measured values are close to each other

what are pilot studies used for

help plan procedures, assess validity, check techniques, appropriate range of values for independent variable, establish number of repeats required to give a representative value for each independent data points

what does a pilot study allow

evaluation and modification of experimental design

what is an independent variable

variable that is changed in a scientific experiment

what is a dependent variable

variable that is measured in a scientific experiment

what is a discrete variable

can have a finite range of values

what is a continuous variables

can take on infinitely many, uncountable values

what is a simple experiment

involves one independent variable

what is a pro and con of a simple experiment

easy to conduct in the field, findings may not be applicable to a wider setting

what is a multifactorial experiment

involve combination of more than one independent variable or combination of treatments

what is an observational study

the independent variable is not directly controlled for ethical or logical reasons

what are observational studies good at detecting

correlation

what is a con of observational studies

does not directly test a hypothesis so are less useful for determining causation

what are confounding variables

other variables besides the independent variable that may affect the dependent variable

how can confounding variables be monitored

be held constant or monitored so their effect on results can be accounted for in analysis

when would a randomised block design be used

where confounding variables cannot be easily controlled

what is a randomised block design

randomised blocks of treatment and control groups distributed in a way that the influence of any confounding variable is likely to be the same across the treatment and control groups

what is a negative control

provides the results of the experiment in absence of treatment

what is a positive control

included to check that the system can detect a positive result when it occurs

what is a placebo

a treatment without the presence of the independent variable

what is the placebo effect

a measurable change in the dependent variable as a result of a patients expectation rather than changes in independent variable

what are in vivo studies

using a whole living organism in an experiment

what are in vitro studies

performing a given procedure in a controlled environment outside of a living organism

what are examples of in vitro studies

cell growing in culture medium, proteins in solution and purified organelles

What are the advantages of in vitro studies

better control of your experiment and any confounding variables

What are the disadvantages of in vitro studies

not possible for many biological studies

what are the advantages of in vivo studies

better idea of what is going on inside the cell or organism

what are the disadvantages of in vivo studies

living organisms show great variability which cannot really be controlled

what dictates the suitable sample size

extent of natural variation within a population

what requires a larger sample size

more variation in a population

what should a representative sample have

share the same mean and the same degree of variation about the mean as the population as a whole

what is random sampling

members of population have an equal chance of selection

what is systematic sampling

members of population are selected at regular intervals

what is stratified sampling

population divided into categories that are sampled proportionally

what might variation in results be due to

reliability of measurement methods, variation in specimens

what can reliability of measuring instruments or procedures be determined by

repeated measurements or readings of an individual datum point which indicates precision of the measurement

what does a narrow range of values indicate

low variation

what is qualitative data

subjective and descriptive

what is quantitative data

measured objectively usually with a numerical value

what is ranked data

numerical values replaced by rank and sorted from lowest to highest

what does overlapping error bars suggest

no significant difference between data sets

what are box plots used for

show variation within and between data sets

when does correlation exist

when there is a relationship between two variables, does NOT imply causation

when does causation exist

if changes in values of the independent variable are known to cause changes to the value of the dependent variable