Principles of Probability and the Normal Distribution Curve
PRINCIPLES OF PROBABILITY
- Research studies begin with a general question about an entire population, but the
actual research is conducted using a sample
- INFERENTIAL STATISTICS: to use sample data as the basis for answering questions about
the population
- This inference is usually done by following the Principles of Probability
- The principles of probability help us determine the possible outcome in a
situation
WHAT IS PROBABILITY?
- The ratio of the number of ways in which a favored event can occur over
the total number of ways in which that event can occur.
Eg.
- In tossing a coin, the probability that you will get heads is 1⁄2; tails is 1⁄2.
- In drawing a diamond card in a deck of playing cards, the probability is 13/52 (13
diamond cards out of 52 cards); drawing a case is 4/52 (4 ace cards out of 52 cards)
There are many statistical procedures related to probability and predictions but one of the
most useful concepts is the NORMAL DISTRIBUTION CURVE (Bell Curve, Gaussian Curve) by
Johann Carl Friedrich Gauss.
THE NORMAL DISTRIBUTION CURVE
For the purpose of the social sciences, it is important to remember that most observed values
follow the normal distribution curve. Considering the central limit theorem, variables are
normally distributed as the mean, media, and mode are found in the center of the curve.
Characteristics of the Normal Distribution Curve (NDC)
*bell - shaped
*symmetrical
*asymptomatic
- Total area of the NDC is expressed in a proportion: 1.00
- Frequencies, number of cases, etc., are formed into proportions under the NDC
- ALWAYS REMEMBER that the MEAN, MEDIAN, and MODE will always be found in the
CENTER (Central Limit Theorem)
In probability, we are dealing with two types of problems: Type I and Type II.
Type I Problem
- from scores, expectations are computed about the frequencies predicting f
Step 1: Transform raw scores into a deviation score or a standard score, namely,
a zscore
𝑥−𝑥
𝑍 = 𝑆𝐷
Step 2: ESTIMATE and draw the NDC and plot whether what you are computing
for is:
A: area from mean to z
B: area in the larger proportion (body)
C: area in the smaller proportion (tail)
Then, check the z-score in the appropriate column of the NDC tables. (proportion (p) of
frequency (f) for the area). This is found in Appendix A (see Appendix A).
Step 3: Obtain f (where p is the proportion):
𝑓 = (𝑝)(𝑁)
**NOTE: We observe the Principle of Inclusiveness (that any point that goes
beyond the whole value is inclusive for the next whole value: for example
obtaining 8.1 is considered for the value of 9). Not rounding off!
Step 4: Draw conclusions.
EXAMPLE: The average science aptitude IQ of 1,800 high school students is 98 with an SD of
15. Assuming that the scores are normally distributed, how many students will have an IQ of
110? (Observe that we are predicting for f)
GIVEN: 𝑁 = 1800, 𝑆𝐷 = 15, 𝑥 = 110, 𝑀𝑒𝑎𝑛(𝑥) = 98
STEP 1:
𝑋−𝑋
𝑍= 𝑆𝐷
110−98
𝑧 = 15
z = .80
STEP 2:
STEP 3:
𝑓 = (𝑝)(𝑁)
𝑓 = (. 7881)(1, 800)
𝑓 = 1418. 58 ≈ 1419
f = 1419
STEP 4:
1419 students with an IQ of 110 in a science aptitude test.
Type II Problem
- from frequencies, expectations are computed about scores we are predicting scores
Step 1: Transform f into a proportion
𝑓
𝑝= 𝑁
Step 2: Draw the NDC and plot whether what you are computing for is:
A: area from mean to z
B: area in the larger proportion (body)
C: area in the smaller proportion (tail)
Check the appropriate column of NDC table, the z-value for the needed
proportion (remember Inclusiveness!)
Step 3: Transform z-scores into raw scores:
𝑥 = 𝑥 + 𝑧(𝑆𝐷)
Step 4: Draw conclusions
EXAMPLE: The top 10 students (from our example in Type I Problem) in the aptitude test will
be given a scholarship. What IQ score must a student at least obtain to qualify for the
scholarship?
GIVEN: 𝑓 = 10, 𝑁 = 1, 800, 𝑥 = 98, 𝑆𝐷 = 15
STEP 1:
𝑓
𝑝 = 𝑁
10
𝑝 = 1800
𝑝 =. 0056
STEP 2:
STEP 3:
𝑥 = 𝑥 + 𝑧(𝑆𝐷)
𝑥 = 98 + 2. 53(15)
𝑥 = 135. 95
STEP 4:
To get a scholarship, IQ must be 135.95 and above.
EXAMPLE PROBLEM (QUESTION 1): How MANY of them have a starting pay ranging from 7,500
Pesos to 10,000 Pesos? This is a Type I Problem.
STEP1: Compute for z-score. You should get -2.5. Though this is negative, we simply look for
+2.5 in the Appendix A.
STEP2:
STEP3: Compute for f. You should get 56 (or 57).
STEP4: Conclusions: ___________ employees are getting paid between 7,500 to 10,000
pesos
EXAMPLE PROBLEM (QUESTION 2): What is the salary RANGE of 50 fresh graduates belonging
to the middle part of the distribution? This is a Type II Problem. However, since it is asking
for a range, our approach will be different.
STEP1: Get the p. You should get .4348. Since we are looking for a range of proportion in the
middle, we divide .4348 into two: that is + - .2174
The corresponding z-value of +- .2174 is +- 0.58 (remember principle of inclusiveness).
STEP2:
STEP3: Compute for x. You should get two scores:
1st computation should be: 10,000 + ((.58)(1000))= 10,580
2nd computation is: 10,000 + ((.58)(1000))= 9,420
STEP4: The range of salary in the middle part is.
