Statistical Tests Flashcards
What is a control groups
A separate group you set up within your experiment where you treat them exactly the same without applying the independent variable
What do you always say when asked how something acts as a control group
“ to compare to “ “to show the effect of…. Then name the indépendant variable “
What should you say when it asks to describe how a control group should be treated
It should be treated with exactly the same conditions as the experimental group
Apart from the indépendant variable shouldn’t be applied to this group
What do you say if the experimental group are being given a tablet or injection of a drug
The control group should be given an injection or tablet as well just without the drug
What should you say if it asked to suggest what the control should be in the experiment
If the experiment involved liquid then the control normally involves using water
If the experiment involved using enzymes the control is normally no enzymes or a boiled enzymes.
What should you say if asked to explain why you use a control group or explain how group x acts as a control
Group X allows you to compare to group y so you can see the effect of (independant variable)
This is to ensure that no other factor is having an effect on the data
What can sometimes happen however
Control groups can sometimes show the effect of more than one factor
It’ll be worth 2 marks and you need to mention both of these factors
What is the common level of probability used in biology
5% written as p= 0.05
What happens if you calculated value is equal to or greater than the critical value for p= 0.05
There is a significant different between X and Y
There is a less than 5% probability than the different js due to chance
Reject the null hypothesis
There is likely to be a biological cause behind the difference
What happens if you calculated value is less than the critical value for p=0.05
There isn’t a significant different between X and Y
There is a greater than 5% probability that the different is due to chance
Accept the null hypothesis
Cannot be confident of a biological cause
What happens if you are given a P value in the question
They have already carried out the stats test and you need to conclude without comparing this number to a critical value
What happens if you are given the number 0.05
Say 5% in your answer
If you were given 0.02 then say 2% if it were 0.001 it’s 0.1%
What about the signs
If it is P= 0.05 then the probability the difference is due to chance IS 5%
If it’s P< 0.05 then it’s LESS THAN 5%
If it’s P> 0.05 then it’s MORE THAN 5%
When do you use the Chi squared test
To compare frequencies in different catagories or comparing expected and observed frequencies in catagories
What is the formula for X2
(See in booklet)
What would be the expected numbers if you want to collect earthworms from field A and B
Do the average so if field A is 167 and field b is 198 do 167+198/2 which is 182.5
What are the observed values in that experiment
167 and 198
In the chi squared table what should you fill in
O value
E vakue
(O-E)
(O-E)2
(O-E)2
Divided by E
What is the weird symbol thingy value
You add up the last column of numbers
What is the degrees of freedom
The number of catagories - 1 so for example
2-1
What is the critical value
For 1 degree of freedom the critical value is 3.841
2 degrees it is 5.991
3 degrees is 7.815
4 degrees is 9.488
5 degrees is 11.071
What happens if the calculated value ( 2.64 is less than 3.841) Is the less than the critical value
the null
There is no significant difference
There is less than a 5 % probability the difference is due to chance
What null hypothesis do you write for the chi squared test
There is no difference in the number of (dependant variable) between group 1, group 2 etc
What is the null hypothesis for the correlation coefficient
There is no correlation between (variable 1 ) and variable 2
When do you use the correlation coefficient
When looking for a correlation between 2 variables
(Normally represented as a scatter graph or line graph)
What is the method for the correlation coefficient
It’s suitable for data that can be ranked from highest to lowest
1) rank each set of data (highest number is 1 next highest is 2 etc)
2) if you have 2 or more pieces of data that are the same number give one a rank eg 5 and one 6 then average it
3) find the difference between 2 ranks for each pair of values
4) square the difference
5) total the square of the differences
What is the equation for the correlation coefficient
See in booklet
What does each number mean
rs- speareman rank correlation coefficient
n- number of pairs of items in the sample (10)
D- difference between each pair of ranked measurements
Substitute numbers into the formula
What does the value for the speareman rank do
It always lies between +1 and -1
So if it is near to +1 it has a positive correlation if it is close to 0 it has no correlation if it is close to -1 it has a negative correlation
What conclusion can you make
As pH increase heights of buttercups decreases if it’s negative
As pH increase heights of buttercups increases if it me positivr
What would the critical value be
If 5 measurements then it’s 1
6= 0.89
7=0.79
8=0.74
9=0.68
10=0.65
When do you use a Chi squared test
Used to compare frequencies of organisms to see whether or not the number of organisms in one catagory is significantly different in another catagory
What is the formula for the Chi squared test
The sum of the observed minus expected value squared divided by expected
How do you work out the expected value
the average between the thingies
What happens if the value is less than the critical value
Accept the null
Theresno significant difference
Theres more than a 5% probability that the difference is due to chance
What happens if the value is more than the critical value
Reject the null
There is a significant difference
Less than 5% probability the difference is due to chance
When do you use the students T test
When comparing 2 means
What is the formula for the students T test
Mean of sample 1 - mean of sample 2
divided by
SD of sample 1/No of sample for sample 1 add SD of sample 2/No of sample for sample 2
What is the method for the Students T test
1) First calculate the mean and standard deviation of each sample
2) Enter numbers into the equation
3) Calculate how many degrees of freedom you have using the calculation n1-1 + n2-1
What is the uncertainty of an instrument
Half the smallest resolution. For example a thermometer is likely to have an uncertainty of +/- 0.5 if the graduation is 1 celsius apart
What is the equation for percentage uncertainty
uncertainty divided by reading times by 100
What do you do when combining uncertainties
You must work out the uncertainties of each separate first then add them together
How do we reduce uncertainty
Choosing instruments withmore graduations at smaller intervals or higher resolutions
Repeating the reading
What is standard deviation
A measure of how psread out your repeats are from your mean
What are the advantages of using the SD instead of the range
It shows the spread of data around the mean whereas the range is only the difference between the highest and lowest value
Reduces the effect of anomalies but the range includes anomalies
can be used to indicate whether a difference between results is significant
How would you describe which statistical test to use
1) Say to write a null hypothesis
2) State which test to use and say to calcualte its test statistic
3) Look up in table to find the critical value of p=0.05
4) Use the figure to accept or reject the null
With correlations can you conclude anything
No because correlation does not equal a casual relationship as it may be due to another factor ( common 2 marker)
How do you answer “ Explain what the results of the statistical test shows”
1) state if the difference in distribution is due to chance or not for any species
2) State why (p>0.0001 which is highly significant)
Why is it useful to express results in ratio form
It allows for comparision as the masses/volumes/populations are different
What should ratios be displayed in the form of in biology
_____:1
How would you calculate the ratio
Divide the first number in the question by the second number
How do you calculate percentage change
Final value- initial value divided by initial value x 100
How do you draw a tangent to find the rate of reaction
Find the point of the graph where you need to know the rate
Angle the ruler so that there is equal space between the ruler and curve
Draw a line between the ruler using a sharp pencil
How do you calculate the pH of a solution
-log10(H+)
How do you calcualte specific heat capacity
Mass x Specific heat capacity x temperature change
What are the conversions from micrometres, millimetres and nanometres
mm-nm=x10^6
nm-mm=divide 10^6
mm to micro- x1000
nano to micro- divide by 1000
milli to micro - x1000
micro to milli - divide by 1000
What is the equation for magnification
Image length divided by actual length
How do you calculate the length of a certain stage of mitosis
Number of cells in a certain stage divided by total number of cells
x100
How do you calcualte mitotic index
number of cells undergoing mitosis
divided by total number of cells
What do you always say when asked to suggest 2 factors to consider
Ethical considerations
Larger number of sample to increase reliability
How do you identify the concentration of an unknown solution of glucose
You use to formula
C1 x V1= C2 X V2
How do you plot a calibration curve using a colorimeter for a substance other than glucose
1) Make a dilution series of known concentrations of ( name the substance you are identifying )
2) Use a colorimeter to measure the absorbance of each concentration
3)Plot a graph of concentration on the x axis and absorbance on the y axis and draw a calibration curve
4) Use the absorbamce of the sample to find the concentration from the calibration curve