What is Probability ?
If we toss a fair coin, there are two possible outcomes, a head (H) or a tail (T); that is N=2. So the P(H) = (0.5). If a toss of two coins, how many possible outcomes ?
First coin
Second coin H T
H T
What is the probability of : - P(2H) - P (at least 1H) -P( 1H and 1T)
A pregnant woman wonders about the chance of having a boy or a girl baby. Vital statistics datas indicate there are about 1056 live births of boys for every 1000 live birth of girls, so she estimate her probability of having a boy as 1056 / 2056 = 0.514
Probability :
Probability applies exclusively to a future event, never to a past (outcome of event is unknown) Probability may be used to measure the uncertainty of the outcome of such events, for example , the probability of surviving to age 80 of developing cancer Probability statements are Numeric, in the range of 0 1 ( 0: the event will not happen, 1 the event will happen with certainty)
Events :
Mutually exclusive event Not mutually exclusive event Complementary event Independence event
Event A
AB Event B
Complementary event
Event Ac is the complement of event A
Ac
For example : 100 patients Lung cancer, 20 patients still a live for 3 years.P(A) = 20/100 = 0.2), so that 80 patients Acdied for 3 year 0.8)1 P(A) )=80/100 = P(
Independent event
Two events are independent if the occurrence of one has no effect on the chance of occurrence of the other. The multiplication rule For example : the outcomes of repeated tosses of a coin, because the outcomes of one toss does not affect the outcomes of any future toss.
Probability Distributions
A key application of probability to statistics is estimating the probabilities that are associated with the occurrence of different events. Help us reach a decision whether certaint events are significant or not Mathematical distributions
Probability Distributions
Variables continue
Normal Distributions Sample Mean Distributions ( t- student distributions) F Distributions
Variables discrete :
Binomial distributions Chi Square distributions
the Normal Distribution Properties : Symmetrical Bell shaped curve extending infinitely in both directions Have area under the curve. The total is 1 It is a theoretical distribution defined by two parameters: and The mean and median of a Normal Distribution are equal
(Kuzma, 2005; p.
200 800 -3 +3
300 -2
400 -1
500
600
700
+1 +2
(Kuzma, 2005; p.
(a)
68%
+ +1,96 +2,58
68%
(b)
-1
+1 +1,96
(Kuzma, 2005; p.
Example 1
What is the proportion of persons having Stat score between 500 and 650 ? X 500, sd=100
Z= Z= x 500 500 = =0 100
Z table is 0.4332
SAT math Z
1,5
120
140
Berapa peluang untuk mendapatkan Tekanan darah antara 100 dan 130 mm Hg ?
0,1587
Z = 1.0 0.1587
example 2 :
We assume that the systolic blood pressure of large group Of adult men are approximately normally distributed, That the mean systolic BP is 120 mmHg, and standard deviation is 10 mmHg. If one adult men is randomly selected What is the probabilty that his systolic BP will be less than 140 mmHg
The t-distribution
Derived by W.S Gossett who published under the pseudonym Student, it is often called Students tdistribution Using Degree of Freedom Its shape similar to that of the Standard Normal Distribution, but it is more spread out longer tails. Its shape approaches Normality as the degrees of freedom increase It is particularly useful for calculating confidence intervals for and testing hypotesis about one or two means We use it when the population SD is not known
x1
n2 n3
x2 x3 xi
Gambar grafik himpunan nilai x ini mempunyai gambar spt distribusi normal
dst
Ciri-ciri distribusi normal dapat diterapkan utk menduga
Probability 0.3
0.2
0.1
-4 4
-2 t
2 (Kuzma, 2005; p.
Degree of Freedom
The sample size minus the number of parameters that have to be estimated to calculate the statistic. They indicate the extent to which the observation are free to vary
Distribusi Binomial
Ciri ciri ;
Bila jumlah n tetap dan p kecil maka distribusi yang dihasilkan akan menceng kanan dan bila p makin besar maka kemiringan akan berkurang dan bila p = 0.5 maka distribusi akan simetris dan bila p > 0.5 maka distribusi yang dihasilkan menceng kiri Bila p tetap dengan jumlah n yang makin besar maka akan dihasilkan distribusi yang mendekati distribusi simetris
5 3,84
10 9,49
15 12,59