Testing a Claim Flashcards
What is validity
How well a method tests a claim
What are factors that can affect the validity of an experiment
Choosing a suitable method
Using accurate measuring equipment correctly
Only change one variable whilst ensuring all other variables are kept the same
Repeating in order to assess reliability and recording an average
What is reliability
The trustworthiness of your results, how close repeated measurements are to each other
What are some things you can do to increase reliability
Repeat the experiment a statistically significant number of times
Identify and remove outliers
Record an average
How is reliability assessed?
By checking for similarity between repeat data. This can be done through standard deviation and standard error
What is a random error
unpredictable fluctuations of results around a mean. They cannot be stopped but can be minimised by taking more data and recording an average
What is a systematic error
Predictable fluctuations of data which result in data that is consistently incorrect. They can be stopped by ensuring that all equipment is calibrated before use and used in accordance with the manufacturers instructions.
False claims example
Ribena marketed itself as a healthy alternative to other fruit drinks because of supposedly high levels of vitamin C. Ultimately it was found that Ribena did not have as much vitamin c as claimed.
Emotive advertising
Using pathos to advertise their product. For the case of Ribena repeated use of the word “glow” was used to describe health benefits. In the advert it showed children playing and colour returning to the screen following the consumption of the drink as well as the growth of plants and general flourishing of the environment.
What is accuracy
How close a measured value is to a true or accepted value
What are factors affecting accuracy
experiment minimises systematic errors
equipment with high resolution and precision ie digital pH logger over universal indicator and the human eye
Equipment is calibrated correctly
Why should you have a large sample size
More closely represent a larger population
Random error in smaller samples has a greater impact on overall results
How is standard error calculated
SE = SD/sqrt(N)
How is the coefficient of variation calculated
CV = SD/mean
What values of CV are considered very reliable, reliable and not reliable
CV<0.1 is very reliable
0.3<CV<0.5 is reliable
CV>0.5 is not reliable
