The Department of Census and Statistics has published the preliminary results of the 2016 Computer Literacy Survey.
The survey has its beginnings in the e Sri Lanka initiative which supported the initial iterations starting from 2004. This is the sixth in the series.
One expects indicators such as literacy and device ownership to increase every year. But not in 2016. Most of the indicators are plateauing or declining (not yet statistically significantly). For example, household ownership of desktop/laptop computers has declined to 22.5 percent from 24.4 percent in 2015, with the sharpest drop experienced in the urban sector (from 41.5 to 35.3 percent). Computer literacy for the country as a whole has increased from 26.6 to 27.5 percent from 2015 to 2016. But the most significant result is that it has declined from 39.9 to 38.5 in the urban sector. This suggests that computer literacy has reached its highest point and that it will show an island-wide decline in 2017 or whenever the next survey is done, if it is done in the same form.
How does this connect to my intentionally provocative question about this being the last of the computer literacy surveys?
It is normal for indicators such as ownership and literacy to plateau. But they plateau after reaching 80 or 90 percent. Here, we are seeing a leveling off or decline well before 50 percent coverage has been achieved.
That suggests only one explanation. The phenomenon has changed. The questions that are being asked are obsolete and no longer capable of capturing changes in the phenomenon.
From our surveys and qualitative research in the region (last time we did a survey in Sri Lanka was in 2011), we know that smartphones are the device used by most people at the bottom of the pyramid to connect to the Internet. Data from Facebook, Google etc confirm that they get most traffic from mobile devices in our countries. The wording in the survey questions misses this transformation completely.
In our surveys (completed in Myanmar in 2016; to be fielded in Pakistan, India, Bangladesh, Nepal, Cambodia and Indonesia in 2017) we formulate the questions very differently, in ways that capture actual use behaviors. It is time the Department also does that. We stand ready to help.
BTW, this is not the first time we’ve said the survey needs to be fixed. See here.