AQA STATISTICS 1 REVISION NOTES
AVERAGES AND MEASURES OF SPREAD
Mode : the most common or most popular data value the only average that can be used for qualitative data not suitable if the data values are very varied
www.mathsbox.org.uk
Mean : important as it uses all the data values Disadvantage affected by extreme values
If the data is grouped use the mid-point of each group as your x
Median : the middle value when the data are in order For n data values the median is the n + 1 th value
2
Not affected by extreme values
For 10 values the median will be the 5 th value halfway between the 5th and the 6th values
Range biggest value smallest value - greatly affected by extreme values Interquartile Range Upper quartile Lower quartile - measures the spread of the middle 50% of the data and is not affected by extreme values
4 LQ 3 LQ
7 M 4 M
9 9 UQ IQR = 9 - 4 = 5 7 UQ 8 9
IQR = 7.5 3.5 = 4
Standard Deviation Deviation from the mean is the difference from a value from the mean value The standard deviation is the average of all of these deviations Formulas to work out standard deviation are given in the
SCALING DATA Addition if you add a to each number in the list of data : New mean = old mean + a New median = old median + a New mode = old mode + a Standard Deviation is UNCHANGED Multiplication - If you multiply each number in the list of data by b : New mean = old mean b New median = old median b New mode = old mode b New Standard Deviation = old standard deviation b
www.mathsbox.org.uk
PROBABILITY
Outcome : each thing that can happen in an experiment Sample Space : list of all the possible values
NOTATION
AB
A and B both happen
A B
C'
either A or B or both happen
C doesnt happen
P(C) = 1 P(C)
P(A B) = P(A) + P(B) P(A B)
Mutually Exclusive Events two or more events that cannot happen at the same time
P(A B) = P(A) + P(B)
Independent Events the outcome of one event does not affect the outcome of another P( A B ) = P(A) P(B) Conditional Probability : when the outcome of the first event affects the outcome of a
second event, the probability of the second event depends on what has happened
P(B/A) means the conditional probability of B given A P(B/A) = P(A
B) so P( AB
)= P(A) P(B/A)
P(A)
q
If the question states that the events are independent then a tree diagram might be a good idea multiply along the branches then add the appropriate combinations together Probability of at least 1 = 1 Probability of none If you are asked to find probabilities using data in a table work out the row/column totals before you start
q q
www.mathsbox.org.uk
BINOMIAL DISTRIBUTION
q
A question is binomial if: Probability of an event happening is given (p) Number of people/trials/objects chosen given (n) EQUALS or EXACTLY use the formula P(x=r) Make sure you write it out with the values substituted in
n C r p (1 p)
from your calculator make sure you write down this value
nr
Check that your powers add to make n (number of trials)
MORE/LESS THAN/AT LEAST use tables Remember tables give less than or equal to Make sure you list the numbers and identify which ones you need to include P(X>5) P(X<5) 0 1 2 3 4 5 6 7 8 9 =1P(X5) 0 1 2 3 4 5 6 7 8=P(X4)
MEAN and VARIANCE Mean = np
Variance = np(1-p)
standard deviation = np(1 p)
ASSUMPTIONS Independent events with a fixed probability of success Randomly selected
COMPARISONS You may be asked to calculate the mean and standard deviation of a binomial distribution and compare them to the mean and standard deviation of a sample (table of results) - if both the means are approximately the same AND both standard deviations (or variances) are approximately equal then you can say that binomial model appears to fit the data and that it must be independent, random observations.
www.mathsbox.org.uk
NORMAL DISTRIBUTION FINDING PROBABILITIES
q
State the mean and variance (standard deviation) Standardise to find the z value
z =
x mean standard deviation
Sketch a graph and shade in the area to represent the probability P(z <1.2)
1.2
P(z >-0.8)
-0.8
Use the table to find the probability - take care with negative z-values P(z > -0.8) = P(z < 0.8)
-0.8
0.8
P(z < -0.8)
1 P(z < 0.8)
-0.8
q
0.8
ALWAYS CHECK YOUR ANSWER WITH YOUR GRAPH if your shaded area is more that and your answer is 0.4 (for example) you know you have gone wrong somewhere!!
(simultaneous equations)
WORKING BACKWARDS to find the mean/standard deviation or both q State the probability you know and sketch a graph P(X < 34) = 0.95 0.05 34
q q
Standard deviation = 8
Use tables to find the appropriate z value Write down the equation used to standardise with all of the known values substituted
1.6445 = 34 mean 8
q
Rearrange to find the mean
www.mathsbox.org.uk
SAMPLES and PROBABILITIES
If you are asked to calculate probabilities involving samples remember to divide the (population) standard deviation by the square root of the sample size when you standardise.
z =
x mean s n
or
z=
x mean variance n
ESTIMATION
estimating the population mean from a sample mean and finding a confidence interval
If a random sample of size n is taken from a normal population and the sample mean x is found, then the 95% confidence interval of the population mean is given by Z value
x 1.96
2
s n
, x + 1.96
s n
where s is either the population variance (if given) or an unbiased estimate of the population variance (found from the sample see below)
CASE 1 : Standard deviation or Variance of the population is stated in the question - from the data you only need to calculate the mean - use your tables to find the appropriate z value - write out the above expressions for the confidence intervals with all your values substituted in - calculate the two values for your confidence intervals and state clearly (3 sf) CHECK your answer add the two answers together and divide by 2 this should be the sample mean!!!
CASE 2 : Standard deviation or Variance of the population is UNKNOWN - you will need to use the data to calculate the variance of the sample and then an unbiased estimate of the population variance - your calculator will give you value of the standard deviation for the sample you have entered
sxn - square this to calculate the sample variance
An unbiased estimate of the population variance is Use this as the value of
q
n Sample variance n1
s in your confidence interval when substituting the values in
INTERPRETATION a 95% confidence interval tells us that if we took the same size sample 100 times then 95 of the confidence intervals we would calculate should contain the TRUE population mean.
www.mathsbox.org.uk
CENTRAL LIMIT THEOREM you only need to use the central limit theorem if you are not told that the sample is selected from a population which is Normally Distributed
The theorem concerns the distribution of the sample means and as long as the sample size is large enough (greater than 30) then the sample means will be normally distributed and so we can calculate confidence intervals
REGRESSION finding the equation of the line of best fit least squares
y = a + bx
Intercept - the value of y when x is zero. e.g. When the temperature is 0C ice cream sales are 10
Gradient the change in y for each unit change in x e.g for every 1 degree rise in temperature sales increase by b
If you are asked to interpret the values of a and b, make sure you discuss it in the context of the question NOT in terms of x and y (see examples above) If you are give a table of values use your calculator to find a and b In your workings state a = . b =.. and show your values substituted into y = a + bx
If you are calculating a and b using the formulae make sure you use the formula book showing how you substituted the values in Always work out b first Use b and the means of x and y to work out a a = (mean of y) - b(mean of x) TO PLOT THE REGRESSION LINE choose 2 different values of x use your equation y=a + bx to work out the predicted y-values - plot the two points and join with a straight line RESIDUAL = OBSERVED(actual value) PREDICTED(using equation y= a + bx)
- the smaller the residuals the greater the accuracy of the line of best fit in predicting values - sometimes an average residual can be used to make predictions using the line of best fit e.g if an individual has an average residual of 5 then to predict for this particular person using the line, 5 should be added to the value predicted using the equation.
RELIABILITY OF PREDICTIONS - Interpolation predicting using an x-value within the range of x-values used to calculate the a and b considered to be a reliable prediction - Extrapolation predicting using an x-value outside of the range of x-values used to calculate a and b UNRELIABLE estimate - because you are assuming that the linear trend continues indefinitely use your common sense to explain why this may be incorrect - watch out for NEGATIVE (or unrealistic) y values which may result for the x values suggested again use your common sense to explain why this is unrealistic
www.mathsbox.org.uk
SCALING either the x or the y values will change the equation
e.g y = 0.5 1.2x If the x values are doubled then the equation becomes y = 0.5 1.2(2x) y = 0.5 2.4x If 5 is added on to each of the y values then the equation becomes y + 5 = 0.5 1.2x y = - 4.5 1.2 x (always rearrange to get y = a + bx)
CORRELATION
The Product Moment Correlation Coefficient : is a numerical measures of the strength and type of correlation denoted by r and will lie in the range -1 r 1
q q
indicates how well the data, when plotted in a scatter graph, fits a straight line pattern NOT APPROPRIATE if the data does not follow a linear pattern when plotted (straight line) so scatter graph is needed to check this If you are give a table of values use your calculator to find r (If you have time its a good idea to check that you have entered your values correctly) If you are calculating using summary values -make sure you use the formula book showing how you substituted the values into the formula
INTERPRETING r make sure you do this in the context of the question (not just positive correlation)
e.g. There appears to be a fairly strong relationship between temperature and ice cream sales, higher temperatures appear to correspond to higher values of ice-cream sales and vice versa.
Scaling data a linear transformation or scaling of one or both of the variables will not affect the correlation coefficient all of the points will stay in the same position RELATIVE to each other
TAKE CARE
q
Not all correlation will be linear
For this data, the correlation coefficient is close to 0 This does not mean that there is no correlation but simply mean that there is no linear correlation (pattern appears to be quadratic)
q
Spurious Correlation A strong correlation between 2 variables does not mean that one thing causes the other high marks in a maths exam do not necessarily cause high marks in a Statistics exam, they are likely to both be dependent on a common third variable : the students mathematical ability Outliers One or two outliers can have a dramatic effect on a correlation coefficient
