Practical Skills Flashcards
What is a prediction/hypothesis?
Specific testable statement about what will happen in experiment
Define precise results
Results that don’t vary much from mean
Define Valid
Free of error
(Valid results answer original question)
How do you obtain valid results?
By controlling all variables to make sure you’re testing thing you want
Define accurate results
Results that really come close to the true value
What can decrease accuracy of results?
Human interpretation of measurement (e.g. determining colour change)
How can precision be reduced?
Reduced by random errors
Define Reproducible
If someone different does experiment, using slightly different method or piece of equipment, results will be the same
Define Repeatable
If same person repeats experiment using same methods and equipment = get same results
Define Calibration
Marking a scale on a measuring instrument
Define Resolution
Smallest change a measuring instrument can detect
Define a zero error
Systematic error caused by using equipment that isn’t zeroed properly
Define a random error
Unpredictable way in which all measurements wary
(e.g. human errors in measuring)
How can you reduce the effect of random errors
By repeat readings & finding the mean
Define a systematic error
Measurement wrong by same amount every time
Define a measurement error
Difference between measured value and true value
Define uncertainty
Amount of error your measurements might have
How can you calculate a percentage error of your measurements?
Name 2 ways you can reduce uncertainty
- Using most sensitive equipment available
- Measure a greater amount of something
Define categoric variables
Values that are labels e.g. names of plants
Define nominal variables
Type of categoric variable where there is no ordering of categories
e.g. red flowers, pink flowers, blue flowers
When is it suitable to use a scatter graph?
When you’re looking at relationship between 2 discrete/independent variables
Name 2 reasons why have a control group with a placebo makes your results more reliable
- Removes researcher biasis
- Control group can’t show psychologial effects
Data is often given as percentages of people dying from each cause.
Explain the advantage of giving these data as percentages. (2)
- Easier to compare if sample size effectively the same
- Different no. of people in each group
If experimental group are given the treatment via injections, suggest how the control group should be treated (2)
- Given only saline
- Otherwise treated exactly the same way
What does standard deviation tell you?
Spread from the mean
Comment on the effectiveness of taxol when used separately and as a combined treatment (related to SD)
SD overlap for OGF with taxol and taxol on its own so not conclusive/could be chance/both treatments effective
Why should you repeat experiments?
- To increase the reliability of your results
- Anomalies can be identified
What does an overlap in standard deviation mean?
Unlikely that any difference (in results) is significant
An investigation was carried out into the effect of carbon dioxide concentration and light intensity on the rate of photosynthesis in a species of plant. The temperature was kept constant during the investigation. Explain why. (2)
- Temperature affects the rate of photosynthesis
- ∴ any change in photosynthesis rate is the result of CO2/light intensity
Explain how the results from tube D help to confirm that the explanations for the other tubes are valid. (1)
Shows that indicator alone doesn’t change colour in light
Explain the advantages of collecting a large number of results (2)
- Easier to spot anomalies/increases reliability of results
- Allows use of statistical test
Explain why both indentical and non-identical twins are used in investigations (2)
- Identical twins show genetic influence/differences
- Non-identical twins also show an environmental/non-genetic influence
Explain why it is an advantage to apply the treatment (i.e. 250 seeds per m2) to each row and each column (2)
- Different envrionment or different variables in field
- Minimises the effect of variables
An investigation to determine whether pH affects the rate of an enzyme controlled reaction. Write a null hypothesis.
There is no significant difference between the rate at which the enzyme works at different pHs
Name 3 statistical tests
- Standard error and 95% confidence limits
- Chi-squared test
- Spearman rank correlation
When should you use standard error and 95% confidence limits?
- When testing for a difference between 2 sets of data
- The data is continuous and means can be calculated
- “looking for significant differences (between mean values)”
When should you use chi-squared test?
- When testing for a difference between 2 sets of data
- The data is in discrete categories
When should you use spearman’s rank correlation test?
When testing for a correlation between 2 sets of data
For correlation coefficient:
Calculated value is _____ than the critcal value so ___ null hypothesis
Calculated value is greater than the critcal value so reject null hypothesis
Calculated value is less than the critcal value so accept null hypothesis
For correlation coefficient:
Why do we reject the null hypothesis when the calculated value is greater than the critical value?
A probability of less than 0.05 or 5% that the correlation in results is due to chance
For correlation coefficient:
Why do we accept the null hypothesis when the calculated value is less than the critical value?
A probability of more than 0.05 or 5% that the correlation in results occurred due to chance
Chi-squared Test:
When do we reject our null hypothesis?
When our calculated value of Chi-squared is greater than the critical value of Chi-squared
Chi-squared Test:
When do we accept our null hypothesis?
When our calculated value of Chi-squared is less than the critical value of Chi-squared
Why do we reject our null hypothesis when our calculated value of Chi-squared is greater than the critical value of Chi-squared?
∵ there’s less than 5% probability that the differences between the observed and expected data are due to chance
Why do we accept our null hypothesis when our calculated value of Chi-squared is less than the critical value of Chi-squared?
∵ there’s more than 5% probability that the differences between the observed and expected data are due to chance
Give the reason why logarithmic scales have been used on the y-axes in the graph
large range of values/numbers
Scientists found a postive correlation between the inhibition of germination and the concentration of the extract. Describe how they could find out whether this correlation was significant. (3)
- Produce null hypothesis
- Carry out Spearman Rank correlation test / find correlation coefficient
- Use values to show P < critical value / find probability of results being due to chance
What does the histogram indicate about the inheritance of this feature? Explain your answer. (2)
- polygenic inheritance / several genes
- many categories / continuous range / single or multiple allele inheritance would produce discrete categories
The standard error of the mean was calculated. What information would this give about the mean height of 17-year-old males? (2)
- (SE gives idea of) variability of mean
- time / population mean would lie within these limits in 68% / 70% / 2 / 3 of samples
Explain why the means and standard deviations are more useful than the ranges for detecting any differences between two samples (3)
- Range = just extreme values / outliers
- OR not typical / not representative / could be anomalies
- Mean and SD uses all the values or less affected by anomalies
- Mean and SD can be used in a statistical test
- OR can be used to see if two results differ significantly
