biostatistics, statistical software iv. statistical estimation, confidence intervals. hypothesis...
TRANSCRIPT
Biostatistics, statistical software IV.
Statistical estimation, confidence intervals. Hypothesis tests. One-and
two sample t-tests.
Krisztina Boda PhD
Department of Medical Informatics, University of Szeged
INTERREG 2Krisztina Boda
Statistical estimation
A parameter is a number that describes the population (its value is not known).
For example: and are parameters of the normal distribution N(,) n, p are parameters of the binomial distribution is parameter of the Poisson distribution
Estimation: based on sample data, we can calculate a number that is an approximation of the corresponding parameter of the population.
A point estimate is a single numerical value used to approximate the corresponding population parameter. For example, the sample mean is an estimation of the population’s
mean, .
approximates
approximates
n
x
n
xxxx
n
ii
n 121 ...
1
)(1
2
n
xxSD
n
ii
INTERREG 3Krisztina Boda
Interval estimate, confidence interval
Interval estimate: a range of values that we think includes the true value of the population parameter (with a given level of certainty) .
Confidence interval: an interval which contains the value of the (unknown) population parameter with high probability.
The higher the probability assigned, the more confident we are that the interval does, in fact, include the true value.
The probability assigned is the confidence level (generally: 0.90, 0.95, 0.99 )
INTERREG 4Krisztina Boda
Interval estimate, confidence interval (cont.)
„high” probability: the probability assigned is the confidence level (generally: 0.90, 0.95, 0.99 ).
„small” probability: the „error” of the estimation (denoted by ) according to the confidence level is
1-0.90=0.1, 1-0.95=0.05, 1-0.99=0.01 The most often used confidence level is
95% (0.95), so the most often used value for is
=0.05
INTERREG 5Krisztina Boda
The confidence interval is based on the concept of repetition of the study under consideration
If the study were to be repeated 100 times, of the 100 resulting 95% confidence intervals, we would expect 95 of these to include the population parameter.
http://www.kuleuven.ac.be/ucs/java/index.htm
INTERREG 6Krisztina Boda
Formula of the confidence interval
for the population’s mean when is known
It can be shown that
is a (1-)100% confidence interval for . u/2 is the /2 critical value of the standard normal distribution, it
can be found in standard normal distribution tablefor =0.05 u/2 =1.96for =0.01 u/2 =2.58
95%CI for the population’s mean
(x un
x un
/ /, )2 2
)96.1,96.1x(n
xn
INTERREG 7Krisztina Boda
The standard error of mean(SE or SEM)
is called the standard error of mean
Meaning: the dispersion of the sample means around the (unknown) population’s mean.
When is unknown, the standard error of mean can be estimated from the sample by:
n
n
SD
nn
SD
n
deviation standard
INTERREG 8Krisztina Boda
Formula of the confidence interval for the
population’s mean when is unknown When is unknown, it can be estimated by the sample SD
(standard deviation). But, if we place the sample SD in the place of , u/2 is no longer valid, it also must be replace by t/2 . So
is a (1-)100 confidence interval for . t/2 is the /2 critical value of the Student's t statistic with n-1
degrees of freedom with n-1 degrees of freedom (see next slide)
),x( 2/2/ n
SDtx
n
SDt
INTERREG 9Krisztina Boda
t-distributions (Student’s t-distributions)
df=19 df=200
INTERREG 10Krisztina Boda
Two-sided alfa df 0.2 0.1 0.05 0.02 0.01 0.001 1 3.078 6.314 12.706 31.821 63.657 636.619 2 1.886 2.920 4.303 6.965 9.925 31.599 3 1.638 2.353 3.182 4.541 5.841 12.924 4 1.533 2.132 2.776 3.747 4.604 8.610 5 1.476 2.015 2.571 3.365 4.032 6.869 6 1.440 1.943 2.447 3.143 3.707 5.959 7 1.415 1.895 2.365 2.998 3.499 5.408 8 1.397 1.860 2.306 2.896 3.355 5.041 9 1.383 1.833 2.262 2.821 3.250 4.781
10 1.372 1.812 2.228 2.764 3.169 4.587 11 1.363 1.796 2.201 2.718 3.106 4.437 12 1.356 1.782 2.179 2.681 3.055 4.318 13 1.350 1.771 2.160 2.650 3.012 4.221 14 1.345 1.761 2.145 2.624 2.977 4.140 15 1.341 1.753 2.131 2.602 2.947 4.073 16 1.337 1.746 2.120 2.583 2.921 4.015 17 1.333 1.740 2.110 2.567 2.898 3.965 18 1.330 1.734 2.101 2.552 2.878 3.922 19 1.328 1.729 2.093 2.539 2.861 3.883 20 1.325 1.725 2.086 2.528 2.845 3.850 21 1.323 1.721 2.080 2.518 2.831 3.819 22 1.321 1.717 2.074 2.508 2.819 3.792 23 1.319 1.714 2.069 2.500 2.807 3.768 24 1.318 1.711 2.064 2.492 2.797 3.745 25 1.316 1.708 2.060 2.485 2.787 3.725 26 1.315 1.706 2.056 2.479 2.779 3.707 27 1.314 1.703 2.052 2.473 2.771 3.690 28 1.313 1.701 2.048 2.467 2.763 3.674 29 1.311 1.699 2.045 2.462 2.756 3.659 30 1.310 1.697 2.042 2.457 2.750 3.646 1.282 1.645 1.960 2.326 2.576 3.291
The Student’s t-distribution
For =0.05 and df=12, the critical value is t/2 =2.179
INTERREG 11Krisztina Boda
Student’s t-distribution
Degrees of freedom: 8
INTERREG 12Krisztina Boda
Student’s t-distribution
Degrees of freedom: 10
INTERREG 13Krisztina Boda
Student’s t-distribution
Degrees of freedom: 20
INTERREG 14Krisztina Boda
Student’s t-distribution
Degrees of freedom: 100
INTERREG 15Krisztina Boda
Student’s t-distribution table
Two sided alfa
Degrees of freedom 0.2 0.1 0.05 0.02 0.01
1 3.077683537 6.313752 12.7062 31.82052 63.65674
2 1.885618083 2.919986 4.302653 6.964557 9.924843
3 1.637744352 2.353363 3.182446 4.540703 5.840909
4 1.533206273 2.131847 2.776445 3.746947 4.604095
5 1.475884037 2.015048 2.570582 3.36493 4.032143
6 1.439755747 1.94318 2.446912 3.142668 3.707428
7 1.414923928 1.894579 2.364624 2.997952 3.499483
8 1.39681531 1.859548 2.306004 2.896459 3.355387
9 1.383028739 1.833113 2.262157 2.821438 3.249836
10 1.372183641 1.812461 2.228139 2.763769 3.169273
11 1.363430318 1.795885 2.200985 2.718079 3.105807
INTERREG 16Krisztina Boda
Student’s t-distribution table
two sided alfa
degrees of freedom 0.2 0.1 0.05 0.02 0.01 0.001
1 3.077683537 6.313752 12.7062 31.82052 63.65674 636.6192
2 1.885618083 2.919986 4.302653 6.964557 9.924843 31.59905
3 1.637744352 2.353363 3.182446 4.540703 5.840909 12.92398
4 1.533206273 2.131847 2.776445 3.746947 4.604095 8.610302
5 1.475884037 2.015048 2.570582 3.36493 4.032143 6.868827
6 1.439755747 1.94318 2.446912 3.142668 3.707428 5.958816
7 1.414923928 1.894579 2.364624 2.997952 3.499483 5.407883
... … … … … … …
100 1.290074761 1.660234 1.983971 2.364217 2.625891 3.390491
... … … … … … …
500 1.283247021 1.647907 1.96472 2.333829 2.585698 3.310091
... … … … … … …
1000000 1.281552411 1.644855 1.959966 2.326352 2.575834 3.290536
INTERREG 17Krisztina Boda
Example 1. We wish to estimate the average number of heartbeats per minute
for a certain population. The mean for a sample of 13 subjects was found to be 90, the standard
deviation of the sample was SD=15.5. Supposed that the population is normally distributed the 95 % confidence interval for :
=0.05, SD=15.5 Degrees of freedom: df=n-1=13 -1=12 t/2 =2.179 The lower limit is
90 – 2.179·15.5/√13=90-2.179 ·4.299=90-9.367=80.6326 The upper limit is
90 + 2.179·15.5/√13=90+2.179 ·4.299=90+9.367=99.367 The 95% confidence interval for the population mean is
(80.63, 99.36) It means that the true (but unknown) population means lies it the
interval (80.63, 99.36) with 0.95 probability. We are 95% confident the true mean lies in that interval.
INTERREG 18Krisztina Boda
Example 2. We wish to estimate the average number of heartbeats per minute for a
certain population. The mean for a sample of 36 subjects was found to be 90, the standard
deviation of the sample was SD=15.5. Supposed that the population is normally distributed the 95 % confidence interval for :
=0.05, SD=15.5 Degrees of freedom: df=n-1=36-1=35 t /2=2.0301 The lower limit is
90 – 2.0301·15.5/√36=90-2.0301 ·2.5833=90-5.2444=84.755 The upper limit is
90 + 2.0301·15.5/√36=90+2.0301 ·2.5833=90+5.2444=95.24 The 95% confidence interval for the population mean is
(84.76, 95.24) It means that the true (but unknown) population means lies it the
interval (84.76, 95.24) with 0.95 probability. We are 95% confident the true mean lies in that interval.
INTERREG 19Krisztina Boda
Example
Descriptives
170.3908 .91329
168.5752
172.2064
170.2886
170.0000
72.566
8.51859
152.00
196.00
44.00
11.0000
.274 .258
.270 .511
Mean
Lower Bound
Upper Bound
95% ConfidenceInterval for Mean
5% Trimmed Mean
Median
Variance
Std. Deviation
Minimum
Maximum
Range
Interquartile Range
Skewness
Kurtosis
Body heightStatistic Std. Error
87N =
Body height
95
% C
I B
od
y h
eig
ht 173
172
171
170
169
168
INTERREG 20Krisztina Boda
Presentation of results
INTERREG 21Krisztina Boda
Hypothesis testing
Hypothesis: a statement about the population
Based on our data (sample) we conclude to the whole phenomenon (population)
We examine whether our result (difference in samples) is greater then the difference caused only by chance.
INTERREG 22Krisztina Boda
Hypothesis
Hypothesis: a statement about the population Examples
H1: =16 (the population mean is 16) H2: ≠16 (the population mean is not 16) H3: B=G (boys and girls score the same on mathematics exams) H4: B≠G (boys and girls score differently on mathematics exams)
Statisticians usually test the hypothesis which tells them what to expect by giving a specific value to work with. They refer to this hypothesis as the null hypothesis and symbolize it as H0. The null hypothesis is often the one that assumes fairness, honesty or equality.
The opposite hypothesis is called alternative hypothesis and is symbolized by Ha. This hypothesis, however, is often the one that is of interest.
INTERREG 23Krisztina Boda
Steps of hypothesis-testing Step 1. State the motivated (alternative) hypothesis Ha. Step 2. State the null hypothesis H0. Step 3. You select the probability of „error”, or the α significance
level. α =0.05 or α =0.01. Step 4. You choose the size n of the random sample Step 5. Select a random sample from the appropriate
population and obtain your data. Step 6. Calculate the decision rule. Step 7. Decision. a) Reject the null hypothesis and claim that your alternative
hypothesis was correct the difference is significant at α100% level.
b) Fail to reject the null hypothesis correct the difference is not significant at α 100% level .
INTERREG 24Krisztina Boda
Testing the mean of a sample drawn from a normal population: one sample t-
test Problem, data.
The normal value of the systolic blood pressure is 120 mm Hg. The following are the systolic blood pressures (mm Hg) of n=9 patients
undergoing drug therapy for hypertension. 182.00 152.00 178.00 157.00 194.00 163.00 144.00 114.00 174.00
Summary statistics The mean=162 mmHg, the standard deviation SD=23.92 .
Question Can we conclude with 95% confidence on the basis of these data that
the population mean is =120? HO: the population mean is 120, =120 Ha: the population mean is not 120 , 120 Assumption: the sample is drawn from a normally distributed
population
INTERREG 25Krisztina Boda
Decision rule based on confidence interval
Find the 95% CI for the above data! α=0.05 mean=162 SD=23.92 The standard error, SE=SD/ n=7.97 t8,0.05=2.306) The confidence interval: (mean - t*SE, mean + t * SE )=(162-2.306*23.92/9,
162+2.306*7.97)=(143.61,180.386) Can we conclude with 95% confidence on the basis of these data that the population
mean is =120? No, because the confidence interval does not contain 120. Decision rule based on confidence interval: is the given number (the number in the
null hypothesis) in the confidence interval? If yes: the difference is not significant at α level If not: the difference is significant at α level
In our case 120 is not in the 95%Ci, the difference is significant at 5% level.
INTERREG 26Krisztina Boda
Decision rule based on t-value Calculate
t= (mean - c)/SE=(162-120)/7.97=5.26. Degrees of freedom: n-1=9-1=8 Compare the absolute value of the calculated
t to the critical t-value in the table: t8,0.05=2.306 5.26>2.306 Decision rule:
if |t|>ttable, the difference is significant at α level if |t|<ttable, the difference is not significant at α
level The acceptance (non-rejection) region is the
set of values for which we accept the null hypothesis (- ttable, ttable)
The critical region (rejection region) is the set of values for which the null hypothesis is rejected.
t=5.26
Acceptance region
INTERREG 27Krisztina Boda
Decision rule based on p-value
The probability, computed assuming that H0 is true, that the test statistic would take a value as extreme or more extreme than that actually observed.
Decision: If p<, then the difference is
significant at level If p>, then the difference is not
significant at level
In our case the difference is significant at 5% level, because p=0.001<0.05
One-Sample Statistics
9 162.0000 23.92175 7.97392VAR00001N Mean Std. Deviation
Std. ErrorMean
One-Sample Test
5.267 8 .001 42.0000 23.6121 60.3879VAR00001t df Sig. (2-tailed)
MeanDifference Lower Upper
95% ConfidenceInterval of the
Difference
Test Value = 120
INTERREG 28Krisztina Boda
One-sample t-test, example
A company produces a 16 ml bottle of some drug (solution). The bottles are filled by an automated bottle-filling process. If this process is substantially overfilling or under filling bottles, then this process must be shut down and readjusted. Overfilling results in lost profits for the company, while under filling is unfair to consumers. For a given adjustment of the bottles consider the infinite population of all the bottle fills that could potentially be produced. We let denote the mean of the infinite population of all the bottle fills.
The company has decided that it will shut down and readjust the process if it can be very certain that the mean fill is above or below the desired 16 ml.
Now suppose that the company observes the following sample of n=6 bottle fills:
15.68, 16.00, 15.61, 15.93, 15.86, 15.72 It can be verified that this sample has mean=15.8 and standard deviation
s=0.156. Question: Is it true that the mean bottle fill in the population is 16?
INTERREG 29Krisztina Boda
1. H0: =c, (c given constant). H0: =16.
2. Ha: c. Ha: 16
3. Select =0.05or =0.01 4. Find the sample size n n=6 5. Select a random sample x x xn1 2, ,..., . 15.68, 16.00, 15.61, 15.93, 15.86, 15.72
6. The decision rule:
Confidence interval
Critical points (t-value)
p-value
(x ts
nx t
s
n / /, )2 2 s
cxn
n
scx
t
and
find t/2
p (Sig. level Sig. 2 tail Prob....)
)96.15,639.15()6
1532.057.215.8,
6
1532.057.2(15.8
197.361532.0
)168.15(
t
t/2 =2.57
2-tail Sig=0.024
INTERREG 30Krisztina Boda
7. Decision: Confidence
interval Critical points
(t-value) p-value
a) Ha : reject H0,
the difference is significant at (1-)100%-level.
the confidence interval does not
contain c
t /2 t
p <
a) H0 : do not reject H0,
the difference is not significant at (1-)100%-level.
the confidence interval contains the
value c
t /2 t
p >
The decision in our example the difference is significant at 95%-level
16 is not in the interval (15.64, 15.96)
|t|> t/2 |-3.197|=3.197>2.57
0.024 < 0.05
INTERREG 31Krisztina Boda
A one-sample t test for paired differences (paired t-test)
Before treatment After treatment Difference x1 y1 d1 x2 y2 d2 . .
.
. . .
xn yn dn H0: =0 (the mean of the population of differences is 0)
Ha: 0 (the mean of the population of differences is not 0)
td
sn
d
, where d is the mean of the sample of differences, sd is the standard deviation of
the sample of differences
INTERREG 32Krisztina Boda
Comparison of the means of two related samples: paired t-test
Self-control experiment (measure the data before and after the treatment on the same patient) or
Related data: Before treatment – after treatment Left side – right side Matched pairs
Null hypothesis: the two sample means are approximations of the same population mean (there is no treatment-effect, the difference is only by chance)
HO: before= after or difference= 0 Alternative hypothesis: there is a treatment effect HA: before≠ after or difference≠ 0
INTERREG 33Krisztina Boda
Comparison of the means of two related samples: paired t-test (cont.)
Calculation: take the difference of the two samples, calculate the mean and SE of the differences
Fix The paired t-test is a one-sample t-test for the
differences. The decision rule is based on: Confidence interval for the mean difference Calculation of t-value and comparison its absolute
value with the ttable
p-value (software)
INTERREG 34Krisztina Boda
Paired t-test, example
A study was conducted to determine weight loss, body composition, etc. in obese women before and after 12 weeks of treatment with a very-low-calorie diet .
We wish to know if these data provide sufficient evidence to allow us to conclude that the treatment is effective in causing weight reduction in obese women.
The mean difference is actually 4. Is it a real difference? Big or small? If the study were to be repeated, would we get the same result or less, even 0?
Before After Difference 85 86 -1
95 90 5 75 72 3110 100 10 81 75 6 92 88 4 83 83 0 94 93 1 88 82 6105 99 6
Mean 90.8 86.8 4.SD 10.79 9.25 3.333
INTERREG 35Krisztina Boda
Paired t-test, example (cont).
Idea: if the treatment is not effective, the mean sample difference is small (close to O), if it is effective, the mean difference is big.
HO: before= after or difference= 0 HA: before≠ after or difference≠ 0 Let =0. Degrees of freedom=10-1=9, t0.05,9=2.262 Mean=4, SD=3.333 SE=3.333/10=1.054
INTERREG 36Krisztina Boda
Paired t-test, example (cont.)
Decision based on confidence interval: 95%CI:(4-2.262*1.054, 4+2.262*1.054)=(1.615,
6.384) If H0 were true, 0 were inside the confidence
interval Now 0 is outside the confidence interval, the
difference is significant at 5% level, the treatment was effective.
The mean loss of body weight was 4 kg, which could be even 6.36 but minimum 1.615, with 95% probability.
INTERREG 37Krisztina Boda
Decision based on test statistic (t-value): t= (mean -
0)/SE=mean/SE=4/1.054=3.795. This t has to be compared to the critical t-value in the table.
|t|=3.795>2.262(=t0.05,9), the difference is significant at 5% level
Decision based on p-value: p=0.004, p<0.05, the
difference is significant at 5% level
Acceptance region
ttable, critical value
tcomputed, test statistic
INTERREG 38Krisztina Boda
Testing the mean of two independent samples from normal populations: two-
sample t-test Independent samples:
Control group, treatment group Male, female Ill, healthy Young, old etc.
Assumptions: Independent samples : x1, x2, …, xn and y1, y2, …, ym the xi-s are distributed as N(µ1, 1) and the yi-s are distributed as
N µ2, 2 ).
H0: 1=2, Ha: 12
INTERREG 39Krisztina Boda
The case when the standard deviations are equal Assumptions:
1. Both populations are approximately normal. 2. The variances of the two populations are equal
( 1= 1 = ). That is the xi-s are distributed as N(µ1,) and the yi-s are distributed as
N(µ2,) H0: 1=2, Ha: 12
If H0 is true, then
has Student’s t distribution with n+m-2 degrees of freedom. • Decision: If |t|>tα,n+m-2, the difference is significant at α level, we reject H0
If |t|<tα,n+m-2, the difference is not significant at α level, we do not reject H0
tx y
sn m
x y
s
nm
n mp
p
1 1 .s
n s m s
n mpx y22 21 1
2
( ) ( )
INTERREG 40Krisztina Boda
The case when the standard deviations are not equal
Both populations are approximately normal. 2. The variances of the two populations are not equal ( 1 1 ). That is the xi-s are distributed as N(µ1, 1) and the yi-s are distributed
as N(µ2, 2) H0: 1=2, Ha: 12
: If H0 is true, then
has Student t distribution with df degrees of freedom. • Decision: If |t|>tα,n+m-2, the difference is significant at α level, we reject H0
If |t|<tα,n+m-2, the difference is not significant at α level, we do not reject H0
dx y
sn
s
mx y
2 2
)1()1()1()1()1(
22
ngmgmn
df g
sn
sn
s
m
x
x y
2
2 2.
INTERREG 41Krisztina Boda
Comparison of the variances of two normal populations: F-test
H0: 1=2
Ha:1 > 2 (one sided test) F: the higher variance divided by the smaller
variance:
Degrees of freedom: 1. Sample size of the nominator-1 2. Sample size of the denominator-1
Decision based on F-table If F>Fα,table, the twp variances are significantly different at α
level
Fs s
s sx y
x y
max( , )
min( , )
2 2
2 2
INTERREG 42Krisztina Boda
Table of the F-distribution α=0.05
számláló-> nevező 1 2 3 4 5 6 7 8 9 10
1 161.4476 199.5 215.7073 224.5832 230.1619 233.986 236.7684 238.8827 240.5433 241.88172 18.51282 19 19.16429 19.24679 19.29641 19.32953 19.35322 19.37099 19.38483 19.39593 10.12796 9.552094 9.276628 9.117182 9.013455 8.940645 8.886743 8.845238 8.8123 8.7855254 7.708647 6.944272 6.591382 6.388233 6.256057 6.163132 6.094211 6.041044 5.998779 5.9643715 6.607891 5.786135 5.409451 5.192168 5.050329 4.950288 4.875872 4.81832 4.772466 4.7350636 5.987378 5.143253 4.757063 4.533677 4.387374 4.283866 4.206658 4.146804 4.099016 4.0599637 5.591448 4.737414 4.346831 4.120312 3.971523 3.865969 3.787044 3.725725 3.676675 3.6365238 5.317655 4.45897 4.066181 3.837853 3.687499 3.58058 3.500464 3.438101 3.38813 3.3471639 5.117355 4.256495 3.862548 3.633089 3.481659 3.373754 3.292746 3.229583 3.178893 3.13728
10 4.964603 4.102821 3.708265 3.47805 3.325835 3.217175 3.135465 3.071658 3.020383 2.978237
Nominator->
Denominator|
INTERREG 43Krisztina Boda
ExampleControl group Treated group 170 120 160 130 150 120 150 130 180 110 170 130 160 140 160 150 130 120 n=8 n=10 x=162.5 y=128 sx=10.351 sy=11.35
sx2=107.14 sy
2=128.88
s
t
p2 7 107 14 9 128 88
10 8 2
749 98 1160
16119 37
162 5 128
119 37
10 8
18
34 5
10 924 444 6 6569
. . ..
.
.
.
.. .
Our computed test statistic t = 6.6569 , the critical value int he table t0.025,16=2.12. As 6.6569>2.12, we reject the null hypothesis and we say that the difference of the two treatment means is significant at 5% level
F 128 88
107 141 2029
.
.. ,
Degrees of freedom 10-1=9, 8-1=7, critical value int he F-table is F,9,7=3.68. As 1.2029<3.68, the two variances are considered to be equal, the difference is not significanr.
INTERREG 44Krisztina Boda
Result of SPSSGroup Statistics
8 162.5000 10.35098 3.65963
10 128.0000 11.35292 3.59011
csoportKontroll
Kezelt
BPN Mean Std. Deviation
Std. ErrorMean
Independent Samples Test
.008 .930 6.657 16 .000 34.50000 5.18260 23.51337 45.48663
6.730 15.669 .000 34.50000 5.12657 23.61347 45.38653
Equal variancesassumed
Equal variancesnot assumed
BPF Sig.
Levene's Test forEquality of Variances
t df Sig. (2-tailed)Mean
DifferenceStd. ErrorDifference Lower Upper
95% ConfidenceInterval of the
Difference
t-test for Equality of Means
INTERREG 45Krisztina Boda
Two sample t-test, example.
A study was conducted to determine weight loss, body composition, etc. in obese women before and after 12 weeks in two groups:
Group I. treatment with a very-low-calorie diet . Group II. no diet Volunteers were randomly assigned to one of these
groups. We wish to know if these data provide sufficient
evidence to allow us to conclude that the treatment is effective in causing weight reduction in obese women compared to no treatment.
INTERREG 46Krisztina Boda
Two sample t-test, cont. Data
Group Patient Change in body weight Diet 1 -1 2 5 3 3 4 10 5 6 6 4 7 0 8 1 9 6 10 6 Mean 4. SD 3.333 No diet 11 2 12 0 13 1 14 0 15 3 16 1 17 5 18 0 19 -2 20 -2 21 3 Mean 1 SD 2.145
INTERREG 47Krisztina Boda
Two sample t-test, example, cont.
HO: diet=control, (the mean change in body weights are the same in populations)
Ha: diet control (the mean change in body weights are different in the populations)
Assumptions: normality (now it cannot be checked because of small
sample size) Equality of variances (check: visually compare the two
standard deviations)
INTERREG 48Krisztina Boda
Two sample t-test, example, cont.
Assuming equal variances, compute the t test- statistic: t==2.477
Degrees of freedom: 10+11-2=19 Critical t-value: t0.05,19=2.093 Comparison and decision:
|t|=2.477>2.093(=t0.05,19), the difference is significant at 5% level p=0.023<0.05 the difference is significant at 5% level
477.2238.5
19
01025.64999.99
3
1110
1110
109
145.2103333.39
14
11 22
mn
nm
s
yx
mns
yxt
pp
INTERREG 49Krisztina Boda
Circulation,2004;109:1212-1214.
Example from the literature
INTERREG 50Krisztina Boda
INTERREG 51Krisztina Boda
Step1. H0, HA Basic Question is: Are these two groups different? Formalized as
H0: p1 = p2 (null, no difference) HA: p1 ≠ p2 (alternative, it is what we want to prove)
or H0: 1 = 2 (null, no difference) HA: 1 ≠ 2 (alternative, it is what we want to prove)
The formalization of H0 and HA depends on the data type and on the measure of the endpoint.
INTERREG 52Krisztina Boda
Step 2. Fix
is generally chosen to be 0.05 It represents the mistake of our later
decision when reject H0.
INTERREG 53Krisztina Boda
Step 3. Decide the sample size n
The sample size depends on… The objective(s) of the research (estimation, hypothesis test, …) Main outcome: categorical or continuous; one or more, primary
or secondary, and the estimation of the distribution of the outcome – based on earlier trials.
Probability of type I. error, - see later The power of the test (1-) (1- probability of type II. error) –see
later The research design effect size of clinical importance …
INTERREG 54Krisztina Boda
Step 4. Collect data (perform the experiment)
INTERREG 55Krisztina Boda
Step 5. Decision rule Test statistic: the summary statistics (signal/noise)
of the data used in decision making between H0 and Ha. Compute the t-statistic. When the null hypothesis is true, the probability
distribution of the test statistic (null distribution) is known .
Based on the null distribution, the critical values can be found according to the given .
. the p-value can be computed (probability of the
observed test statistic as is or more extreme in either direction when the null hypothesis is true).
INTERREG 56Krisztina Boda
Step 4. Decision.
Decision based on p-value If p<α: statistical significance at α level. H0 is
rejected in favour of Ha.
If p≥ α : non-significance at α level. H0 cannot be rejected.
Decision based on test statistic compare the computed t-value to the critical t-value. The
difference is significant, if the absolute value of the test statistics > critical t-value
Decision based on confidence interval the confidence interval does not contain 0
INTERREG 57Krisztina Boda
Compare the mean age in the two groups! The sample means are „similar”. Is this small difference really caused by chance?
INTERREG 58Krisztina Boda
Step 1. H0: the means in the two populations are equal: 1=2 HA: the means in the two populations are not equal: 1≠2
Step 2. Let α=0.05
Step 3. Decision rule: two-sample t-test.
Step 4. Decision. Decision based on test statistic:
Compute the test statistics: t=-1.059, the degrees of freedom is 14+13-2=25
ttable=2.059 |t|=1.059<2.059, the difference is not significant at 5% level.
p=0.28, p>0.05, the difference is not significant at 5% level.
INTERREG 59Krisztina Boda
How to get the p-value?
If H0 is true, the computed test statistic has a t-distribution with 25 degrees of freedom.
Then with 95% probability, the t-value lies in the „acceptance region”
Check it: now t=-1.0590.025 0.0250.95
ttable, critical value
INTERREG 60Krisztina Boda
How to get the p-value? If H0 is true, the computed test
statistic has a t-distribution with 25 degrees of freedom
Then with 95% probability, the t-value lies in the „acceptance region”
Check it: now t=-1.059 The p-value is the shaded
area, p=0.28. The probability of the observed test statistic as is or more extreme in either direction when the null hypothesis is true.
0.025 0.0250.95
ttable, critical value tcomputed, test statistic
INTERREG 61Krisztina Boda
Review questions and exercises
Problems to be solved by hand-calculations ..\Handouts\Problems hand IV.DOC, ..\Handouts\Practice_t-test.doc
Solutions ..\Handouts\Problems hand IV solutions.DOC, ..\Handouts\Practice_t-test, solutions.doc
Problems to be solved using computer ..\Handouts\PRACT4M.DOC
INTERREG 62Krisztina Boda
Main tests to compare two groups Comparing means, normality supposed (t-tests)
Paired samples (before-after treatment, left side-right side, etc.): paired t-test
Independent samples (control-treatment, treatment 1 – treatment 2, etc): two-sample t-test
Comparing means medians, normality not supposed : nonparametric tests, e.g. tests based on ranks Paired samples : Wilcoxon test Independent samples Mann-Whitney U-test
Comparing „percentages” (frequencies): Paired samples : chi-squared test Independent samples: McNemar test