Measuring the Dynamics of Information Societies: Empowering Stakeholders Amid the Digital Divide

Abstract

Accurate insight into the emergence of information societies is essential not only for understanding the social effects of information and communication technologies, but also for empowering stakeholders to promptly and appropriately respond to the challenges they encounter. One much-discussed challenge that is particularly in need of analytical clarity is the digital divide, which is difficult to empirically elaborate, given its complicated nature. It is prone to superficial interpretations that suit particular agendas. To address this problem, this article proposes a methodology that integrates and upgrades the analysis of absolute change, relative change, and time distance into a general multidimensional approach. With this methodology, target audiences have an intuitively persuasive and methodologically sound instrument that could reinforce trust in digital divide studies. The approach is applied in evaluating the Internet penetration gap between Slovenia and Denmark, which often serves as a benchmark for policymaking in Solvenia.

Keywords:

• digital divide measurement
• absolute differences, relative differences
• time distance
• the diffusion process, statistical literacy
• policy assessment

APPENDIX: THE BASIC THREE STATISTICAL MEASURES FOR THE PRESENTED SCENARIOS

Table

Notes

¹For more information, see also http://www.gaptimer.eu/overview_of_the_methodology.html

²OECD questionnaire (OECD 2009): “Does any member of this household/do you have access to the Internet at home regardless of whether it is used?«; GSS questionnaire (translated): “Does your household have access to the Internet at home?”; Eurostat (SORS): “Do you or anyone in your household have access to the Internet at home, regardless of whether it is used?”

³Statistics Denmark also provides information on information societies but its unit of observation is the family. The interesting effect is that it reports higher penetration rates than does Eurostat (e.g., for 2009, Statistics Denmark 86% and Eurostat 83% of households with access to the Internet).

⁴Third-order polynomial trend lines; R²_DK = 0.9935 and R²_SI = 0.9964.

⁵The bases for this conclusion for Slovenia are the data provided by RIS (a national project on the use of the Internet in Slovenia, http://www.ris.org) and for Denmark are the data from the Denmark Statistical Yearbook 2010 (http://www.dst.dk/HomeUK/Statistics/ofs/Publications/Yearbook/2010.aspx).

⁶Both distribution curves and corresponding annual changes in penetration are approximated with simple third-order polynomial functions (as mentioned in notes 4 and 8 and in Table 1). These calculations can be further modeled with general time series (e.g., ARIMA) or more specific diffusion models.

⁷Denmark: y = 0.0127x³ − 0.7269x² + 14.51x − 11.556; R² = 0.998; Slovenia: y = 0.0009x³ − 0.1864x² + 7.3053x − 7.0426; R² = 0.9959.

⁸An increase in the time distance in the last phase of Internet adoption is expected because of the asymptotic narrowing of the difference between the two functions.

⁹Very different time horizons are used in forecasting, usually classified as short-term, mid-term, and long-term forecasts. For this scenario, a time frame of 21 years is taken (2009–2030). Considering other classifications of forecasts’ time horizons (see Albright 2002; Weingand 1995), this time frame fits in long-term predictions. However, the »real« time horizon is long (13 years), which may lead to more accurate forecasts.

¹⁰The corresponding polynomial equations are: Denmark: y = 0.0068x³ − 0.4914x² + 11.985x − 5.6095; R² = 0.9912; Slovenia: y = 0.0025x³ − 0.2071x² + 7.3111x − 6.8056; R² = 0.9981.

¹¹In Scenario A, the growth rates from 2010 forward are on average 1.18 (a 10-year average) in Denmark and 1.61 in Slovenia. In Scenario B, the 20-year averages for Slovenia and Denmark are 2.23 and 0.89, respectively (due to very low growth rates in the last 10 years for almost absolute penetration). The 10-year average growth rates for Slovenia and Denmark are 2.76 and 1.21, respectively.

¹²The corresponding polynomial equations are: Denmark: y = 0.0122x³ − 0.6889x² + 14.094x − 10.695; R² = 0.9978; and Slovenia: y = 0.0032x³ − 0.1787x² + 6.7697x − 5.4631; R² = 0.9961.

¹³Of course, we should differentiate between purposeful one-sided reporting of results based on only one statistical measure that shows findings in line with the reporter's interests on one hand, and goal-oriented use of statistics that implies a well-thought choice of particular statistical measure or method on the other hand. If the choice of the measure or method can be methodologically justified and if the purpose and potential impact of the study are appropriately taken under consideration, the usage of that particular measure or method should not be considered as negative (for more details, see Hilbert 2011).