You are on page 1of 10

YEAR 6 TYPE II PORTFOLIO 2013

Anglo-Chinese School (Independent)

POPULATION TRENDS IN CHINA

I undersigned, hereby declare that the following course work is all my own work and that I worked independently on it __________________

Name: Class: Index Number: Subject Teacher:

Ng Wei Xuan 6.14 Joppa 16 Ms Audrey Tay


Page 1 of 10

YEAR 6 TYPE II PORTFOLIO 2013

Introduction
In this portfolio, I will investigate the different functions that best model the population of China from 1950 to 1995. The table below represents the population of China: Year Population in Millions 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995

554.8 609.0 657.5 729.2 830.7 927.8 998.9 1070.0 1155.3 1220.5

Using the graphing software Graphmatica, the above information can be represented in a data plot by plotting the population (in millions) against the year.

Comments: The population of China from 1950 to 1995 was generally increasing at an increasing rate. This data plot could either be represented by an exponential graph, quadratic graph, or a cubic graph.

Page 2 of 10

YEAR 6 TYPE II PORTFOLIO 2013

Cubic graph
For cubic graphs, the standard equation is ( )( )( ) , where

(Population in Mil) t (year) Amplitude of the graph Horizontal translation Vertical translation

First attempt: Using trial and error, I substituted random values into that fits the data plots as closely as possible. Plot ( )( )( ) and to get a curve

Comments: However, the curve does not fit the data plots as it only passes through 3 of them. In order for the curve to fit, it has to be shifted to the right by approximately 20 units, and upwards by approximately 200 units, so that the inflexion of the curve cuts the data plots.

Page 3 of 10

YEAR 6 TYPE II PORTFOLIO 2013 Second attempt: By tweaking the equation using trial and error, the second curve is able to fit most of the data plots. Plot ( )( )( )

Comments: The curve is able to pass through all of the data plots rather accurately, and comes extremely close to the points that it does not pass through. This means that the population trend in China could be represented by this cubic equation.

Limitations of the model: While the population may increase from years 1935 to 2005, there is a gradual decrease from 2005 onwards. There population continues to decrease until it is zero when it is the year 2037 and then negative from then on. As it is impossible for the population of any country to be zero, much less a negative value, it is not possible that the population trend be represented by a cubic equation. That being said, there is still a possibility that this cubic equation could hold true for all the years before 2037, as long as there is still a positive value for the population in China.

***

Page 4 of 10

YEAR 6 TYPE II PORTFOLIO 2013 A researcher suggests that the population, P, at time t can be modeled by

( )

where K, L and M are parameters.

Using the graphic display calculator, the values of K, L and M can be found through logistic regression by 1) Entering the figures into the spreadsheet

2) Using logistic regression function to obtain a suitable curve and to determine the unknown values of K, L and M

The equation obtained defined the parameters as: K = 1946.18370756 L = 4.33655524274 1028 M= 0.033321365681

Page 5 of 10

YEAR 6 TYPE II PORTFOLIO 2013 Hence, the model that the researcher was suggesting is

( )

By plotting the model

( )

and the data plots on

the same axes, the following graph is obtained:

Comments: The curve obtained from the model passes through most of the points and comes extremely close to those it doesnt pass through. Hence, it can be said that this logistic model proposed by the researcher is an accurate representation of the population trend in China from 1950 to 1995.

Page 6 of 10

YEAR 6 TYPE II PORTFOLIO 2013 In order to use the above model to predict future trends of the population in China, the x-axis is extended. The following graph is obtained:

Comments: The graph initially increases at an increasing rate, but plateaus as it nears 2000. Assuming this logistic model holds true for the population trend in China, China is expected to have an increasing population until the year 2100. From 2100 onwards, the population in China remains constant.

Limitations of the model: However, if this model were to hold true, it would mean that the population in China was extremely close to zero in the 18th and 19th century, as shown in the graph below:

As it is virtually impossible for the human population in China to be so sparse in recent history, the curve would not hold true for values during this period of time. This would be one limitation of the logistic model.

***
Page 7 of 10

YEAR 6 TYPE II PORTFOLIO 2013 Therefore, in order to determine which of the two models would be a better representation of the statistics, I shall use additional data published by the International Monetary Fund (IMF), represented in the table below to compare the models. Year Population in Millions 1983 1030.1 1992 1171.7 1997 1236.3 2000 1267.4 2003 1292.3 2005 1307.6 2008 1327.7

*** Firstly, the additional data plots are plotted on the same axes as the original data plot. The general trend is obtained:

Comments: The general trend of the population continues to be increasing. However, from the given data plots, the population seems to be increasing at a decreasing rate from the year 1983 onwards. At this point in time, both curve would seem to fit the above data plots.

Page 8 of 10

YEAR 6 TYPE II PORTFOLIO 2013 Secondly, the two curves are plotted with the new set of data plots and then compared:

Where the pink curve represents the cubic model and the black curve represents the logistic model.

Comments: Comparing the two curves, the logistic curve would fit the points better, although not completely. It passes through 3 out of the 7 plots and comes relatively closer to the other 4 plots, as compared to the cubic curve. The cubic curve passes through only 1 of the plots and does not come as close to the rest of the plots as compared to the logistic curve. Hence, it can be concluded that the logistic curve would fit the new data plots better.

Page 9 of 10

YEAR 6 TYPE II PORTFOLIO 2013 Therefore, to allow the logistic curve to fit the all data plots better, it is modified such that the curve fits the all data plots better. This is done through a series of trial and error.

The modified equation is ( )

Comments: The modified curve is almost a perfect fit for the entire of data plots. It passes through all the points and can be said to be an accurate representation of the population trends in China.

Conclusion
In conclusion, while the first cubic curve might have been a suitable fit, there are limitations that make the model applicable only for a short period of time. It only fits the first set of data and not the second set. Furthermore, due to the nature of the cubic curve, it experiences a negative gradient after the inflexion. Hence, it would be impossible to use this model to predict future population trends. On the other hand, the modified logistic model would be a better and more accurate representation of the population trend, as it passes through the all the data plots. It can also be used to predict the future population trends in China. It can be concluded that the population trend in China is best represented by the logistic model ( )

Page 10 of 10

You might also like