Statistical Consultants Ltd |
      |
ICT Development Index 2010Data Tables, Statistical Modelling Date posted: The International Telecommunication Union (ITU) is a United Nations agency responsible for information and communication technology (ICT). The ITU have published several ICT related indices, including an ICT Development Index (IDI) and an ICT Price Basket (IPB) for most countries. The IDI is a composite of 11 indicators, and is used to compare the overall level of ICT development between countries. The IDI has three sub-indices based on ICT access, use and skills. The IPB is a composite basket based on the prices for fixed telephone, mobile cell-phone, and fixed broadband internet services, expressed as a percentage of average income levels. The indices were first published by the ITU in 2009, using 2008 data. At the time of writing, the most recent versions were published in 2011, and based on 2010 data. ICT Development Index (IDI) by Country
IDI versus GDP per capita (PPP) IDI increases at a decreasing rate with respect to GDP per capita (PPP). The GDP per capita (PPP) figures were from the World Bank, mostly from 2010 (with a few from earlier years). A regression model was fitted to quantify this relationship. Regression model output: Call: lm(formula = log(IDI2010) ~ log(GDPpcppp2010)) Residuals: Min 1Q Median 3Q Max -0.65459 -0.09432 0.02275 0.13098 0.68191 Coefficients: Estimate Std. Error t value Pr(>|t|) (Intercept) -2.73862 0.12613 -21.71 <2e-16 *** log(GDPpcppp2010) 0.44233 0.01383 31.98 <2e-16 *** --- Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1 Residual standard error: 0.2092 on 148 degrees of freedom (2 observations deleted due to missingness) Multiple R-squared: 0.8735, Adjusted R-squared: 0.8727 F-statistic: 1022 on 1 and 148 DF, p-value: < 2.2e-16 For those who aren’t familiar with regression output: The high (close to one) R-squared statistic shows that the model fits strongly to the data: Multiple R-squared: 0.8735 The very low (close to zero) p-values, show that the parameters of the model are highly significant: Pr(>|t|) <2e-16 *** <2e-16 *** These figures are the coefficient estimates, used to construct the model. Estimate (Intercept) -2.73862 log(GDPpcppp2010) 0.44233 The fitted model is represented by the following equation:  Where: x is GDP per capita (PPP) y is the IDI The fitted regression model curve added to the plot:  The two variables are interdependent i.e. they both depend on each other rather than one being dependent and the other being independent. As GDP per capita (PPP) increases, on average the more affordable the development of information and communication technology becomes. Investing in information and communication technology can result in higher GDP per capita (PPP), due to higher productivity, lower costs of production, and the development of high-tech industries. IDI versus IPB IDI decreases at a decreasing rate with respect to IPB. A regression model was fitted to quantify this relationship. Regression model output: Call: lm(formula = log(IDI2010) ~ log(ICTpricebasket2010)) Residuals: Min 1Q Median 3Q Max -0.71750 -0.12492 0.02376 0.15908 0.49380 Coefficients: Estimate Std. Error t value Pr(>|t|) (Intercept) 1.80628 0.02599 69.50 <2e-16 *** log(ICTpricebasket2010) -0.36628 0.01267 -28.92 <2e-16 *** --- Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1 Residual standard error: 0.2221 on 143 degrees of freedom (7 observations deleted due to missingness) Multiple R-squared: 0.854, Adjusted R-squared: 0.8529 F-statistic: 836.2 on 1 and 143 DF, p-value: < 2.2e-16 For those who aren’t familiar with regression output: The high (close to one) R-squared statistic shows that the model fits strongly to the data: Multiple R-squared: 0.854 The very low (close to zero) p-values, show that the parameters of the model are highly significant: Pr(>|t|) <2e-16 *** <2e-16 *** These figures are the coefficient estimates, used to construct the model. Estimate (Intercept) 1.80628 log(GDPpcppp2010) -0.36628 The fitted model is represented by the following equation:  Where: x is the IPB y is the IDI The fitted regression model curve added to the plot:  The two variables are interdependent. The lower the price of information and communication technologies (relative to income), the more affordable it is to adopt those technologies, resulting in a higher quantity of demand for them (resulting in a higher IDI). The development of information and communication technology infrastructure results in an increase in supply, resulting in lower prices for information and communication technologies. IDI Combined ModelA regression model was constructed which includes both IPB and GDP per capita (PPP). The model has the following structure:  where: y is the IDI x1 is GDP per capita (PPP) x2 is the IPB Regression model output: Call: lm(formula = log(IDI2010) ~ log(GDPpcppp2010) + log(ICTpricebasket2010)) Residuals: Min 1Q Median 3Q Max -0.63013 -0.11554 0.01074 0.12034 0.54496 Coefficients: Estimate Std. Error t value Pr(>|t|) (Intercept) -0.97586 0.38346 -2.545 0.0120 * log(GDPpcppp2010) 0.27327 0.03755 7.277 2.23e-11 *** log(ICTpricebasket2010) -0.15858 0.03139 -5.052 1.34e-06 *** --- Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1 Residual standard error: 0.1862 on 140 degrees of freedom (9 observations deleted due to missingness) Multiple R-squared: 0.8984, Adjusted R-squared: 0.8969 F-statistic: 618.7 on 2 and 140 DF, p-value: < 2.2e-16 The fitted model is represented by the following equation:  where: y is the IDI x1 is GDP per capita (PPP) x2 is the IPB The combined model is similar to the constant price elasticity demand model used in econometrics. A constant price elasticity demand model would have the logged quantity (or logged expenditure) explained by the logged price and other variables that explain demand, including a variable related to income. The coefficient of the logged price explanatory variable is interpreted as the price elasticity of demand i.e. a 1% increase in price, would result in a percentage change in the quantity demanded on average equal to the coefficient estimate. Download the Data SetThe data set containing the IDI, IPB and GDP per capita (PPP) data can be downloaded here. Note that:
External link:ITU Measuring the Information SocietySee also:Regression |
|
|
| Copyright © Statistical Consultants Ltd 2012 |