PART a article reading – tech for good PART b descriptive statistics analysis

PART B: Descriptive Statistics Analysisa Analysis of 3 measurements: Based on the analysis of 3 measurements below, it can be concluded that paid apps receive higher ratings from custome

RMIT UNIVERSITY VIETNAM

2019

ASSIGNMENT COVER PAGE

Nguyen Tran Thao Uyen

S3804819@rmit.edu.vn

Part B: 536 words

Table of Contents

PART A: Article Reading – Tech for Good 2

PART B: Descriptive Statistics Analysis 9

a) Analysis of 3 measurements: 9

I Measurements of Central Tendency 9

II Measurements of Variations 9

III Box-and-Whisker Plot Analysis 10

b) Two other numerical variables 11

1 Active app users 11

2 App retention 11

References 12

PART A: Article Reading – Tech for Good

2

Nguyen Tran Thao Uyen – s3804819

According to the MGI report (2019), the nature of technology has no specific good or bad effects on human or the economy It depends on how technology is designed and applied for specific purpose, which will consequently generate positive or negative outcomes

In terms of economy, the application of advanced technology has both potential economic benefits and risks The potential economic opportunities of technology application mentioned in this MGI report are growth in GDP, increased productivity and income opportunities Meanwhile, wage gap and temporary unemployment risks are major negative economic implications of technology reported in this paper However, it is also stated that technology adoption can tackle the difficulties

of unemployment mentioned earlier.

To begin with, one of the benefits that technology has on the economy is encouraging economic growth Specifically, as reported in MGI paper, technological development has reduced working hours since 1800, which consequently led to growth in life expectancy and world GDP (figure 1) Moreover, one simulation conducted by the MGI suggests that

AI adoption has the ability to raise as much as $13 trillion in the global GDP by 2030, and boost the GDP to grow 1% more annually (Bughin et al 2019)

Figure 1 Reproduced from Bughin et al 2019.

Figure 1 shows that alongside with the development of technology throughout the history, the GDP growth rate constantly increases Until 2010 – the age of digitalization, AI and IoT, world’s GDP has experienced a significant increase by 23 times Specifically, automation and networking in production in 1950 – 1960 fuelled the dramatic increase in GDP in the following periods

Secondly, raising productivity is another positive influence of technology, which is closely associated by growth in wage and employment (Figure 2) This positive change results in higher prosperity – to be more specific, improved productivity from application of technology raises people’s income, hence buying power of people grows, leading to increase in demand for more goods and services, which results in growth of demand for more labour.

Figure 2 Reproduced from Bughin et al 2019.

Figure 2 illustrates an upward trend in the UK’s growth in productivity, which went hand-in-hand with higher employment rate This finding is derived from a research conducted by the MGI to study the effect of technology-driven productivity on employment in the UK from 1760 to 2016

Raising income and working opportunities in technological fields is one significant benefit of technology adoption Digital business models can raise incomes through better technological innovation Connectivity platforms such as eBay and Etsy allow people to earn additional income with lower costs than traditional retail channels Moreover, digital platforms can be

used by independent workers to earn income (figure 3) It is reported that online talent platforms could facilitate up to 60 million people find jobs that are more suitable for their skills or desires, and reduce the cost

of managing human resources by up

to 7% (Bughin et al 2019)

Figure 3 Reproduced from Bughin et al 2019.

Figure 3 above shows the number of people in the United States and 5 countries from the

EU that utilizes digital platforms to earn income 24 million out of all independent workers use digital platforms, and 15% of those share that they have earned income from these modern tools Specifically, up to 63% of independent workers who make a living by selling products use digital platforms as their source of earning income

On the other hand, wage gap is mentioned in this MGI report as one negative impact of technology adoption on the economy Recently, digitization and automation have contributed to huge income gap between high-skill and low-skill workers and pressurize the middle class (figure 4)

Figure 4 Reproduced from Bughin et al 2019.

Figure 4 illustrates the increase in low-wage workers (except for Luxembourg and Finland) and high-wage working professionals in 17 countries while the number of middle-wage jobs decrease significantly, indicating a huge income inequality among the population Since 1900s, lower-wage occupations have seen an increase of approximately 2% to 7%, while high-wage’s increase is up to 10-14% in some countries Low-wage occupations are laborers and service workers whose wages are rarely enough to support their daily needs, let alone saving for retirement

6

Nguyen Tran Thao Uyen – s3804819

Unemployment risk is considered another possible negative economic impact of technology The employment ratio is reported to have reduced by about 0.18% to 0.34% every time one robot is deployed, thus heavily affecting low-skill workers in manufacturing industries that are largely automatized these days (Bughin et al 2019) However, this risk is only temporary for about 5 years before training and rehiring take place Technology adoption opens more doors for new occupations, for example software engineering Technology is also a source for improving training and education, improving the flexibility of labour market.

Figure 5 Reproduced from Bughin et al 2019.

Figure 5 shows that the proportion of unemployment is lower due to higher flexibility in labour-market and on-site training, fuelled by technological tools The top 20% (top quintile) of the population only see 4.56% of unemployment rate in both categories

In conclusion, technology adoption can have both positive and negative economic impacts, according the MGI report Positive areas in which technology adoption positively affects are GDP growth and improved productivity Besides, technology also imposes some potential risks on the economy, including income inequality and unemployment However, according to this report, unemployment can be tackled by implementation of technology into training and education It also depends on the approaches taken by the government, businesses and other stakeholders to tackle the challenges of technological risks, in order to make the most out of technology for sustainable growth

PART B: Descriptive Statistics Analysis

a) Analysis of 3 measurements:

Based on the analysis of 3 measurements below, it can be concluded that paid apps

receive higher ratings from customers than free apps

I Measurements of Central Tendency

Paid Educational Apps >,<,= Free Educational Apps

Figure 6 Measurements of Central Tendency of the Free and Paid Educational Apps Customer

Ratings

Analysis:

In this case, Median is the most suitable among three measurements because customer

ratings are ordinal data – data which includes rankings, and the average ratings of both

apps are the same (mean = 4.25)

The median of paid apps (4.5) is higher than free apps (4.4), indicating half of the

customers give paid apps higher scores than free apps This suggests that paid apps receive

higher preference from the customers when compared to free apps

II Measurements of Variation

Paid Educational

>,<,= Free Educational

Interquartile Range –

IQR

Coefficient of

Variance – CV (%)

8

Nguyen Tran Thao Uyen – s3804819

Figure 7 Measurements of Variation of Paid and Free Educational Apps Customer Ratings

Analysis

The average (mean) customer ratings of both apps are the same, so there is no need to use CV as

CV is typically suitable for situations where the average values are extremely different Therefore, standard deviation – S.D is the most suitable measurement in this case.

Paid apps have higher S.D than free apps, suggesting that paid apps receive extremely high scores

or low scores from customers more frequently than free apps However, it is unknown whether there are more exceptionally high ratings than low ratings or not, so it is inconclusive about which apps are more favored by customers if only using measurements of variation.

III Box-and-Whisker Plot Analysis

Figure 8 Box-and-whisker plots of Paid and Free Educational Apps Customer Ratings

Left side >,<,= Right side Result

Median to Extreme value 3.5 > 0.5 Left-skewed

Figure 9 Summary of box-and-whisker plot of Paid Educational Apps customer ratings.

Left side >,<,= Right side Result

Median to Extreme value 3.4 > 0.6 Left-skewed

Figure 10 Summary of box-and-whisker plot of Free Educational Apps customer ratings.

9

Nguyen Tran Thao Uyen – s3804819

As can be seen from figure 8, 9 and 10 above, both apps have left-skewed distribution, meaning that most customer ratings of both concentrate on the right side where there are higher values However, there are still some differences:

The box plot of paid apps is in higher position than free apps (4.2 – 4.7 compared to 4.0 – 4.7), which shows that 75% of the customers rate paid apps higher than 4.2 while free apps receive 75% of the ratings no less than 4.0

Moreover, the right-side box of paid apps is smaller than free apps, indicating the paid apps receive high scores (higher than 4.5 – paid apps’ center rating value) more frequently from 25% - 50% of the customers than free apps

* Conclusion:

Overall, paid apps receive higher customer ratings than free apps because the analysis of two out of three measurements support this conclusion Specifically, median of paid apps is higher than free apps, indicating higher preference of customers for paid apps In terms of standard deviation (S.D), paid apps have higher S.D, but it is indecisive about which apps have higher ratings due to lack of data Regarding box-and-whisker plots, the size and position of paid apps’ box compared to that of free apps show that they receive higher customer satisfaction as analyzed above

b) Two other numerical variables

1 Active app users

It is suggested that customer preference for an app can be measured based on the number of users who actually use the apps on a regular basis rather than only downloading it This metric is believed to provide data about the number of key customers who show a high level of preference for the app (Sela 2019)

2 App retention

According to Armour (2018), retention refers to the percentage of users that return to an app within 3 months from their first using period (session) This method is believed to be useful when comparing between apps in terms of customer preference, since it shows whether an app is appealing enough for the customers to go back to using it (Sela 2019)

Armour, B 2018, ‘5 Methods For Increasing App Engagement & User Retention’,

Clearbridge Mobile, viewed 16 November 2019, < https://clearbridgemobile.com/5-methods-for-increasing-app-engagement-user-retention/>

Bughin, J, Hazan, E, Allas, T, Hjartar, K, Manyika, J, Sjatil, PE & Shigina, I 2019, Tech for Good: Smoothing disruption, improving well-being, McKinsey Global Institute.

Sela, J 2019, ‘The Best Metrics & Tools for Measuring User Engagement’, Appsee Blog,

blog post, 7 January, viewed 16 November 2019, < https://blog.appsee.com/the-best-metrics-and-tools-for-measuring-user-engagement/>

11

Nguyen Tran Thao Uyen – s3804819