Nature, Properties and Applications of The Normal Curve

THE NORMAL CURVE
HISTORY OF THE NORMAL DISTRIBUTION

The normal distribution was originally studied by Abraham DeMoivre in 1733. He was
curious about its use in predicting the probabilities in gambling. He discovered the formula for
the normal curve as the limiting case characterizing an infinite number of independent events.
In the beginning of the 19th century, two mathematicians Pierre-Simon (the Marquis de
Laplace) and Carl Friedrich Gauss rediscovered the normal curve independently. Gauss was
interested in problems in astronomy, and found the errors of observation to be normally
distributed. The normal curve is also known as Gaussian curve in honour of the work done by
him.
The first person to apply the normal distribution to social data was Adolphe Quetelet (in the
middle part of the 19th century). He created the basic methods of physical anthropometrics, and
believed that the normal curve could be extended to apply to problems in anthropology and
human behavior. He collected data on the chest measurements of Scottish soldiers, and the
heights of French soldiers, and found that they were normally distributed. His conclusion was
that the mean was nature’s ideal, and data on either side of the mean were a deviation from
nature’s ideal. Although his conclusion is arguable, he nonetheless represented normal
distribution in a real-life setting.
In the late 19th century, Sir Francis Galton began the first serious study of individual
differences and found that many mental and physical traits conformed reasonably well to the
normal curve. He gave the normal distribution a central role in psychological theory, especially
the theory of mental abilities.
NATURE OF THE NORMAL CURVE
The normal probability curve, or simply the normal curve, is an extremely important and a
commonly occurring shape for population distributions in natural and behavioral sciences. The
abscissa, or horizontal axis, represents different possible values of X, while the ordinate, or
vertical axis, is referred to as the density and is related to the frequency or probability of
occurrence of X. The assumption of normal curve is of great value in psychological
measurement and in many useful procedures in inferential statistics. A variety of psychological
test scores have been found to approximately follow a normal distribution. It is the most widely
used distribution in statistics as it functions as a model for a distribution of sample statistics and
for making statements of probability.
The normal curve is not a single curve; rather it is an infinite number of possible curves, all
described by the following equation:
N 2 2
Y= e − ( X − µ ) / 2σ
σ 2π
where Y = height of the curve directly above any given value of X
N = number of cases
X = raw score
µ = mean
σ = standard deviation
e = 2.7183
π = 3.1416
It is a mathematical abstraction and is defined by two parameters of the normal curve equation,
μ and σ. For each different μ andσ, we will have a different normal distribution. Thus, the
equation of the normal curve describes a family of normal curves each differing in its mean and
standard deviation.
Variations of normal curves in means and standard deviations
All the members of the family of normal curves, although different, have a number of
properties in common:
Properties of the Normal Curve

1. The normal curve is a bilaterally symmetrical curve. That is, if we were to fold the
curve at its highest point at the centre, we would create two halves, each the mirror image
of the other, with equal areas on both sides of the curve.
2. The normal curve is unimodal i.e. it has only one peak or point of maximum frequency.
3. In a normal distribution curve, the mean, median and mode have the same value. (Mo =
Mdn = Mean). A perpendicular line drawn from the highest point meets the X-axis as a
score where the three values will coincide.
4. In a normal curve, most of the data values tend to cluster around the mean. The further a
data point is from the mean, the less likely it is to occur. Thus scores in the middle have a
greater probability of occurrence than the scores farther away.
5. The shape of the normal distribution reminds us of a bell and it is thus widely known as
the “bell-shaped curve”.
6. From the rounded central peak of the normal distribution, the curve falls off gradually at
both tails, extending indefinitely in either direction and getting closer and closer to the
baseline without actually touching it. It is therefore said to be an asymptotic curve with
respect to the horizontal axis.
7. Although the normal curve goes on indefinitely, the area under the graph is considered to
have a unit of 1.00. This property, in addition to the property of symmetry, implies that
the area in each half of the distribution is 0.50.
8. The normal curve is also known as a mesokurtic curve. Kurtosis refers to the peakedness
or flatness of a curve. A curve more peaked than the normal is said to be leptokurtic and
the one flatter than the normal is called platykurtic. The standard normal curve is neither
too peaked nor too flat and is therefore known as a mesokurtic curve. For the normal
curve, Ku=0.263.
9. In a normal distribution, the mean equals the median exactly and therefore the skewness
is zero (Sk = 0). The more nearly the distribution approaches the normal form, the closer
together are the mean and the median and lesser the skewness.
10. The form of the normal probability curve changes from convexity to concavity at points
known as the “points of inflection”. These inflection points of the curve occur at one
standard deviation above and below the mean i.e. ± 1σ
11. Normal curve is a continuous distribution that describes infinity of observations made
on a continuous scale of measurement. However, recorded observations are discrete and
finite and therefore the normal curve can only be an approximation when used as a model
for events in the real world.
12. There are many possible normal curves and in order to make use of its wonderful
properties, we standardize it by converting every normal distribution to a standard
normal distribution. This is done by transforming each raw score to a z-score. The
resulting standard normal distribution has a mean of 0 and a standard deviation of 1.
13. Although the exact shape for the normal distribution is defined by an equation, the
normal curve can also be described by a constant proportion of the total area that lies
between the mean and any given distance from the mean as measured in standard
deviation units. In a normal curve,
Area between µ± 1σ = .6826

Area between µ± 2σ = .9544
Area between µ± 3σ = .9974 where σ = 1 in terms of z-scores.
Most of the area is covered between ± 3σ. Area left at the extremes is .0026 i.e. .0013
(0.13%) on either side that is not very significant. Some other important values for the
area of the normal curve are as follows:
Area between µ± 1.96σ = .9500
Area between µ± 2.58σ = .9900
Area between µ± .6745σ = .5000
Area between mean and 1σ = .3413
Area between 1σ and 2σ= .1359
Area between 2σ and 3σ= .0214
14. In the normal curve, the first and the third quartile points that contain the middle 50% of
the scores correspond to −0.6745σ and +.6745σ respectively. Thus, the quartile deviation
(Q) or semi-interquartile range in a standard normal distribution (also called probable
error) is 0.6745 (Q=0.6745).
With regard to psychological and educational data, the normal curve is, in many instances, a
very good description of the distribution we shall get if we collect many observations and cast
them into a frequency distribution.

99.74%

APPLICATIONS OF THE NORMAL CURVE
Normal curve as a model for real variables
Since it is a probability distribution, the normal curve is a theoretical ideal. Phenomena such as
height, IQ, memory for random numbers etc., do seem to approximate the theoretical normal
distribution. Because so many phenomena have this characteristic – because it occurs so
frequently in nature – researchers in many fields have made extensive use of normal curve by
applying it to data that they collect and analyze.
But it should be noted that some phenomena in social science simply do not conform to the
theoretical notion of the normal distribution. Many distributions are skewed; others have more
than one peak; some are symmetrical but not bell-shaped. Some variables that tend to show at
least a degree of skewness are size of income, reaction time, frequency of accidents etc. Thus
normal curve cannot be used as a model for all the distributions encountered by the researcher,
but must be used with a great deal of discretion.
Further, the normal curve describes infinity of observations that are on a continuous scale of
measurement. However, recorded observations are discrete and finite and therefore the model is
used only as an approximation.
Normal curve as a model for sampling distributions
The normal curve functions as a model for a distribution of sample statistics e.g. if we draw a
very great number of random samples from a population, each sample of the same size and
compute the mean of each sample, the distribution of this large number of means will
approximate the normal curve. This is a property of utmost importance in inferential statistics.

Nature, Properties and Applications of The Normal Curve

Uploaded by

Copyright:

Available Formats

Nature, Properties and Applications of The Normal Curve

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Nature, Properties and Applications of The Normal Curve

Uploaded by

Copyright:

Available Formats

THE NORMAL CURVE

HISTORY OF THE NORMAL DISTRIBUTION

Variations of normal curves in means and standard deviations

Properties of the Normal Curve

Area between µ± 1σ = .6826

You might also like