ECONOMYNEXT — The Institute for Health Policy (IHP) which has been conducting opinion polls on Sri Lanka’s presidential election said it had to substantially revise its earlier model due to an apparent social desirability effect that had led it to underestimate support for one candidate, though the same effect could also potentially lead to an overestimation.
IHP Executive Director Dr Ravi Rannan-Eliya was quoted as saying in a statement from the institute on Wednesday September 18 that it was forced to revise the model it had used in previous months in order to tackle a substantial rise since July 2024 of respondents over-reporting voting for National People’s Power (NPP) leader Anura Kumara Dissanayake in 2019.
“This apparent social desirability effect led the previous model to underestimate support for Dissanayake, as the model assumed that only 3 percent of the electorate cast such a vote. At the same time, this social desirability effect may be associated with respondents telling interviewers that they will vote for Dissanayake when they do not plan to. IHP cautions that these two effects work in opposite directions, and it is not technically possible at the current time to know with certainty which effect is more important. Despite this, IHP has revised its MRP model to assume that many respondents who report voting for Dissanayake in 2019 did not do so,” the institute said.
The social-desirability effect is a response bias that makes respondents answer survey questions in a way that will be viewed favourably by the surveyors or others. This can take the form of over-reporting “socially acceptable” or under-reporting “socially unacceptable” responses.
According to IHP’s latest findings, Dissanayake would’ve won on a second preference count if Sri Lanka’s presidential election was held in early September. The poll also showed a sharp decline in support for opposition leader Sajith Premadasa and a plateauing for President Ranil Wickremesinghe.
IHP’s Sri Lanka Opinion Tracker Survey (SLOTS) Multilevel Regression and Poststratification (MRP) estimates using interviews conducted up to 13 September showed that Dissayanake led with 48 percent, followed by Premadasa on 25 percent, Wickremesinghe on 20 percent, and Sri Lanka Podujana Peramuna (SLPP) candidate Namal Rajapaksa on 5 percent. These estimates were associated with a margin of error of 3–6 percent for the three leading candidates, IHP said in its statement.
The latest IHP MRP modelling indicates that there was a strong surge in favour of Dissanayake since July, when he briefly trailed Premadasa in voter preferences. These MRP estimates are based on a substantially revised MRP model (version 4.0) that is still in beta testing. Data came from 20,714 interviews conducted from 31 August 2021 to 13 September 2024, including 541 interviews conducted in September 2024.
“With the revised model, IHP estimates that the general trends in voting preference during 2024 were similar to what IHP has previously reported, but that its previous estimates for Dissanayake’s support may have been underestimated by 3–4%.
“Dr Rannan-Eliya commented that in almost all the MRP simulations that were done, Dissanayake had a sufficient lead on the first preference votes to achieve a win once second preference choices were accounted for. He also noted that in 7 percent of the simulations, Wickremesinghe came second overall.” (Colombo/Sep19/2024)