Math internal assessment

Page count:

Mean years of education enrolled in, in relation to unemployment percentages

Candidate code: Hzq414

Table of contents

Criterion A: Presentation : 2

Aim and Approach : 3 - 4

Χ² test for independence: 5 - 8

Pearson's product moment correlation coefficient: 9 - 11

Conclusion and Evaluation: 12 - 13

Works cited: 14

Appendix 15 - 18

1
Criterion A: Presentation

When learning about macroeconomics objectives in my higher level economics class and through

personal research. I found the content regarding unemployment compelling due to its depth.

Unemployment is something I come across regularly, when travelling around the world and it is

something which is universally recognized as unwished-for. There is a great disparity between youth

unemployment rates in different countries. For example, I lived in Singapore where there was an

estimated youth unemployment rate of 8.43% in 2020 ("unemployment rate").and in contrast, my

mother's homeland, Kenya has a youth unemployment rate of an estimated 35% ("Tackling youth"). For

my exploration, I wanted to apply mathematics to a real life situation. In this case, unemployment

because it is a global issue. I will focus on whether there is some sort of correspondence and

relationship between the mean years of education enrolled in and the youth unemployment rates.

Unemployment occurs when individuals are prepared and able to work and actively seeking for

employment, but are unable to find work. The contagious disease, Covid-19 which is characterized as an

epidemic occurring worldwide, led to over 495 million full time equivalent job losses worldwide("Impact

on workers"). When an individual is unemployed for a long while, the loss of income and increased

indebtedness, can lead to personal and social consequences. Personal problems include a sense of

unfulfilled human potential which affects an individual's sense of self esteem. This can lead to great

psychological stress which can affect an individuals level of health. Social consequences include absolute

poverty and homelessness, the deprivation leads to desperation which can potentially lead to higher

crime rates within a country.

2
Aim and Approach:

As I briefly established in the introduction, the aim of this exploration is to determine whether there is

some sort of relationship between mean years of education enrollment and the unemployment

percentage rates. In order to determine whether there is a significant dependence between the two sets

of data variables, I will analyze a specific age and gender group, and the same given year for both sets of

data variables. The mathematical data will analyze the average number of years of school attended by

males in the age-group of 15-24 in the given year of 2014. This data will include primary, secondary and

tertiary education. The second set of data will be males in the age-group of 15-24 who have been

registered as unemployed, in the given year of 2014. The process of my mathematical analysis of data

collected is obtained from an organization called Gapminder Foundation, and the sets of data can be

found in the appendix. In order to determine whether the 2 sets of variables are dependent or not, on

one another, I will perform a Χ² test for independence and Spearman's rank correlation coefficient.

Carrying out a Χ² test for independence will reveal whether the two sets of data are independent of

each other or not. Pearson's product moment correlation coefficient will reveal the extent of which

unemployment increases or decreases as the mean years of education enrolled in, increased (i.e the

strength of the correlation. Performing these tests will demonstrate my understanding of the data,

hence make it more easily conveyable and accessible to read.

Unemployment rate formula

unemployed people
* 100
Labor force

Unemployment rate formula calculates the number of unemployed people divided by the labour force,

i.e the employed, self-employed and those actively seeking work (unemployed).

3
I hypothesize that there will be a correspondence at an extent which is not significant enough to

represent dependence between the two variables, because I predict that the mean number of years of

education enrolled in by males in the is one of many factors leading to the % of males who have been

registered unemployed. Although High school graduate is more likely to be employed and have more job

opportunities available, in comparison to a student dropout. A lot of fields of work do not require

specific degrees and skills which are developed and awarded for the completion of specific courses.

You can experience a job which requires no degrees degree, and still gain a sense of identity and

satisfaction.

The main goal of this exploration is to apply course content into an exploration. I will engage by using

appropriate ICT tools to enhance my mathematical communication skills, in order to develop a logical

and coherent exploration, which is easy to follow and understand. Low unemployment is a high-priority

real life issue

4
Χ² test for independence

(Oxford 387-397)

The first test conducted is a nonparametric test called χ² (Chi-square) of independence. This test

will prove whether there is an association between the categorical variables. I.e, whether the mean

years of education enrolled in is dependent or independent to the percentage registered unemployed.

Using the raw data, which can be found in the appendix. I categorized the data into a table, which would

allow me to perform this test.

Total observed values

Unemployment 0 - 0.15 0.15 - 0.30 0.30 - 0.45 > 0.45 Grand total
percentages

Mean years

>12 18 13 3 2 36

8 - 12 23 18 6 4 51

4-8 6 1 0 0 7

0-4 2 0 0 0 2

Grand total 49 32 9 6 96

5
Expected values (Predicted amount of unemployment percentages based on probability law) all data will

be rounded to three significant figures

Unemployment 0 - 0.15 0.15 - 0.30 0.30 - 0.45 > 0.45 Grand total
percentages

Mean years

>12 (36*49)/
96 (36*32)/96 (36*9)/96 (36*6)/96 36

8 - 12 (49*51)/
96 (32*51)/96 (9*51)/96 (6*51)/96 51

4-8 (49*7)/96 (32*7)96 (9*7)/96 (6*7)/96 7

0-4 (49*2)/96 (32*2)/96 (9*2)/96 (6*2)/96 2

Grand total 49 32 9 6 96

6
Calculations for the following formula step by step. fo = observed (observed are the

set of original values. fe = expected (expected are the second set of calculated values based on the

observed, and ∑ = sum of all final values.

Observed Expected Obs-Exp (Obs-Exp)2 (Obs-Exp)2 /Exp

18 18.375 -0.375 0.140625 0.0076530612

13 12 1 1 0.0833333333

3 3.375 -0.375 0.140625 0.416666667

2 2.25 -0.25 0.0625 0.0277777778

23 26.03125 -3.03125 9.188476563 0.3529786915

18 17 1 1 0.0588235294

6 4.78125 1.21875 1.485351563 0.3106617647

4 3.1875 0.8125 0.66015625 0.2071078431

6 3.572916667 2.427083333 5.890733505 1.648718415

1 2.333333333 -1.333333333 1.777777777 0.7619047616

0 0.65625 -0.65625 0.4306640625 0.65625

0 0.4375 -0.4375 0.19140625 0.4375

2 1.020833333 0.979166667 0.9587673618 0.9392006812

0 0.6666666667 -0.6666666667 0.4444444445 0.6666666667

0 0.1875 -0.1875 0.03515625 0.1875

0 0.125 -0.125 0.015625 0.125

Σ = 6.8877431925

The χ² statistic = 6.89 (Σ = 6.8877431925)

7
In a row * column contingency table, the formula for degrees of freedom (df)

is V = (rows - 1) (columns - 1). In this case r=4 and c=4, therefore: V = (4-1) * (3-1) = 3*2= 6

df= 6

By observing the table of critical values of the χ² distribution. The critical value for 9 degrees of freedom,

set to a significance level of 5% is = 12.952, and the χ² statistic value is = 6.8877431925.

If the χ² test statistics is less than the critical value then I accept the initial hypothesis (I.e variables are

independent). If the χ² test statistic is greater than the critical value I accept the alternative hypothesis

(I.e variables are dependent.)

6.89 < 12.952⇒ Accept initial hypothesis

This conclusion shows that the mean years of education enrolled in, is independent to the

unemployment [Link] correspondence is at an extent which is not significant enough to

represent dependence between the sets of variables.

8
Pearson's product moment correlation coefficient

(Oxford 380-386)

Having performed the χ² test, I came to the conclusion that sets of variables are independent of one

another. In order to determine whether there is a correlation, furthermore the strength of the linear

relationship between mean years of education enrollment and unemployment, I calculated a Pearson's

product moment correlation coefficient

S xy
r=
Sx S y

∑ xy ∑ x ∑ y
S xy= −
n n n

S x = √❑

S y = √❑

∑ = the “sum” of

r = correlation coefficient

In order to apply these formulas into calculating the correlation coefficient, I first had to find

the standard deviation of X = mean years of education enrollment, and Y = Unemployment

percentages

9
 Calculations for Standard Deviation

All figures were found using excel

S x = √❑

∑ x ²=12497.7737

n = 96

❑
x ² = 125.3770079

S x = √❑

S x = 2.192745949

S y = √❑

∑ y ²=5.0140

n = 96

❑
y ² = 0.036899549

S y = 0.123813063

After finding the covariance and the standard deviations for both variables. I proceeded to solve

the correlation coefficient.

10
 ∑ xy ∑ x ∑ y
S xy = −
n n n

∑xy = 209.520958

n=96

∑x = 1074.93

∑y = 18.199

209.520958 1074.93 18.199

S xy= −
96 96 96

S xy = 2.182509979 - (11.1971875 * 0.189572917)

= 0.059826486

S xy
r =S S
x y

S xy= 0.059826486

S x = 2.192745949

S y = 0.12381306

0.059826486
r = (2.192745949)(0.12381306)

r = 0.220363018
The results show a weak positive correlation.
Conclusion & Evaluation:

11
Unemployment rates are one of the most widely reported measures of economic activity. These rates

are utilized as a measure of economic performance. A limitation is that it is complicated to calculate and

obtain an accurate and precise measurement of unemployment percentages because some individuals

escape from the official statistics. Consequently, the official figures available online are an

underestimation of the exact rate of unemployment. The underestimations emerge for the following

reasons. Those who are not willing to work are excluded from the calculations. Secondly there is no

distinction between full-time and part time workers. Involuntary part time workers who cannot find full

time employment are still counted, despite them being inadequately employed. Furthermore, the

unemployment figures make no distinction between the type of work done (for example if an individual

who is highly trained and qualified, works as a waiter, this is still counted as full employment)t. Those

who have retired early although would rather be working, are not included as part of the unemployment

figures, and the same goes with individuals on retraining programmes. Given the complications to

calculate and obtain an accurate and precise measurement of unemployment percentages, the

correlation coefficient from the PPMC and extent of correspondence representing independent

association from the chi-squared, are not 100% accurate. (Hoang 74).

Another limitation is that for both sets of data for the categorical variables, the data was made up of 96

countries. Given that there are around 195 countries worldwide, conducting tests with each individual

country worldwide would definitely prove a greater or less degree of relationship which could

potentially alter the answer to the research question. Now when I reflect, I have come to the conclusion

that II could overcome this limitation to some extent by testing sets of data from a year different to

2014, where more data for more countries is recorded and available.

12
Learning about math models and then applying them to a real life issue was very compelling. According

to the results of the tests conducted the Chi squared revealed that the mean years of education

enrollment and unemployment percentages are independent to one another given the year of 2014 and

gender and age group tested being males within 15-24 years of age, which aligns my hypothesis. Yet the

Pearson's product moment correlation coefficient revealed a weak positive correlation. Perhaps

conducting more tests in the future would reveal a more accurate dependence or independence,

overall.

The process for carrying out the specific tests, helped me strengthen my understanding of the data and

make it more conveyable and accessible to read. Most importantly, I was able to apply course content

into this exploration and use mathematical communication skills to answer the research question. The

use of secondary data could indicate biases which could affect the development of a logical and

coherent exploration, which is easy to follow and understand, and therefore I would use more advanced

websites in the future to decrease the extent of uncertainty.

13
Works cited

"unemployment rate for individuals aged 15 to 24 years in Singapore from 2011 to

2020." Statista, 20 Aug. 2020, [Link]/statistics/708329/
singapore-youth-unemployment-rate/. Accessed 20 Jan. 2021.

"Tackling youth unemployment in Kenya through public-private collaboration."

Business Call To Action, [Link]/news/
tackling-youth-unemployment-kenya-through-public-private-collaboration.
Accessed 20 Jan. 2021.

"Impact on workers of COVID-19 is 'catastrophic': ILO." UN News, 23 Sept. 2020,

[Link]/en/story/2020/09/1073242. Accessed 20 Jan. 2021.

Oxford. Mathematics: Applications and Interpretation. Oxford, Oxford University

Press, 2019.

Hoang, Paul. Economics for the IB DIPLOMA. London, Hodder education, 2014.

"Download the data data." Gapminder, [Link]/data/.

14
Appendix

All data extracted from Gapminder

Unemployme
country Mean years country nt
Albania 11 Albania 0.425
Argentina 11.6 Argentina 0.167
Armenia 11.5 Armenia 0.306
Australia 14.6 Australia 0.141
Austria 11.8 Austria 0.106
Azerbaijan 12.2 Azerbaijan 0.115
Barbados 11 Barbados 0.296
Belgium 13 Belgium 0.24
Bhutan 7.38 Bhutan 0.0858
Bosnia and Bosnia and
Herzegovina 10.8 Herzegovina 0.61
Brazil 8.08 Brazil 0.134
Brunei 15.1 Brunei 0.235
Bulgaria 11.3 Bulgaria 0.238
Burkina Faso 3.68 Burkina Faso 0.0529
Burundi 5.44 Burundi 0.0441
Canada 14.2 Canada 0.15
Chile 12.5 Chile 0.149
Colombia 9.96 Colombia 0.139
Costa Rica 9.8 Costa Rica 0.208
Croatia 10.4 Croatia 0.449
Cyprus 13.2 Cyprus 0.374
Czech Czech
Republic 12.5 Republic 0.15
Denmark 12.8 Denmark 0.137

15
Ecuador 10.1 Ecuador 0.0749
Egypt 10.5 Egypt 0.272
Estonia 10.5 Estonia 0.193
Fiji 11.7 Fiji 0.139
Finland 12.4 Finland 0.228
France 12.4 France 0.251
Germany 12.5 Germany 0.0835
Greece 12.8 Greece 0.474
Guatemala 7.76 Guatemala 0.0458
Honduras 8.15 Honduras 0.0756
Hungary 12.6 Hungary 0.2
Iceland 11.3 Iceland 0.129
Indonesia 9.7 Indonesia 0.155
Iran 11.9 Iran 0.213
Ireland 13.6 Ireland 0.266
Israel 14.2 Israel 0.101
Italy 13.1 Italy 0.413
Jamaica 12.2 Jamaica 0.271
Japan 14.6 Japan 0.071
Kyrgyz Kyrgyz
Republic 12.3 Republic 0.12
Latvia 11.6 Latvia 0.194
Lithuania 11.4 Lithuania 0.196
Luxembourg 11.8 Luxembourg 0.261
Malaysia 12 Malaysia 0.0863
Mali 3.67 Mali 0.0894
Malta 10.6 Malta 0.137
Mauritius 10.7 Mauritius 0.194
Mexico 10.7 Mexico 0.0874
Moldova 12 Moldova 0.0955

16
Mongolia 9.72 Mongolia 0.119
Montenegro 12.1 Montenegro 0.36
Namibia 8.63 Namibia 0.356
Netherlands 13.3 Netherlands 0.124
New Zealand 14.1 New Zealand 0.143
North North
Macedonia 11.7 Macedonia 0.52
Norway 14 Norway 0.0913
Pakistan 7.6 Pakistan 0.0403
Palestine 11 Palestine 0.399
Panama 11 Panama 0.0595
Paraguay 9.02 Paraguay 0.083
Peru 11.3 Peru 0.075
Philippines 10.3 Philippines 0.0836
Poland 12.8 Poland 0.227
Portugal 10.2 Portugal 0.342
Qatar 8.68 Qatar 0.0034
Romania 12.3 Romania 0.236
Russia 12.8 Russia 0.134
Samoa 12.8 Samoa 0.156
Saudi Arabia 11.7 Saudi Arabia 0.213
Serbia 11.2 Serbia 0.461
Seychelles 10.7 Seychelles 0.125
Sierra Leone 5.84 Sierra Leone 0.148
Singapore 13.1 Singapore 0.0623
Slovak Slovak
Republic 11.5 Republic 0.295
Slovenia 11.8 Slovenia 0.194
South Africa 10.7 South Africa 0.48
South Korea 13.8 South Korea 0.112

17
Spain 12.4 Spain 0.534
Sri Lanka 11.2 Sri Lanka 0.156
Suriname 9.25 Suriname 0.0776
Sweden 14.1 Sweden 0.242
Switzerland 11.4 Switzerland 0.0864
Tanzania 7.54 Tanzania 0.0314
Thailand 10.7 Thailand 0.0262
Trinidad and Trinidad and
Tobago 13 Tobago 0.0354
Turkey 10.4 Turkey 0.166
Ukraine 13.1 Ukraine 0.237
United United
Kingdom 13.8 Kingdom 0.189
United States 13.3 United States 0.145
Uruguay 10.3 Uruguay 0.161
Vietnam 9.76 Vietnam 0.0504
Yemen 7.97 Yemen 0.235
Zimbabwe 10.4 Zimbabwe 0.116

18