Likely Voter Methodology

During the pre-election season, the Siena College Research Institute (SCRI) uses a Likely Voter Methodology. This methodology differs from polls of registered voters. The Likely Voter Methodology uses derived information from the voter registration file and recorded information from questions asked during the interview to develop a score of a respondent’s likeness to vote. 

1- Voting History From Voter Registration File

During the pre-election season, the Siena College Research Institute collects a sample from a voter registration list (L-2 Votermapping) consisting of the following parameters:

1A – Voting History of Previously Registered Voters

Registered voters who voted in at least one of the four previous elections as determined by their voter registration file. 

  • Statewide Local General Election*
    • (e.g., If upcoming election is 2022, then 2021 local election)
  • Federal Statewide General Election
    • (e.g., If upcoming election is 2022, then 2020 Presidential election)
  • Statewide Local General Election*
    • (e.g., If upcoming election is 2022, then 2019 local election)
  • State Statewide General Election
    • (e.g., If upcoming election is 2022, then 2018 Gubernatorial election)

*If the voter registration file for a geographic area does not contain the local election data, then only the two general elections are extracted and included in the sample file.

1B – New Voter Registration After Last Election

Registered voters who registered to vote in the geographic area after the most recent general election.

2 – Stated Likeliness to Vote

During the pre-election season, the Siena College Research Institute (SCRI) uses a multi-stage likely voter screen consisting of two questions. All registered voters are asked the questions to generate an index score based on the surveyed voter’s responses. Question #1, is asked near the beginning of the survey, and question #2 is asked near the end of the survey.

2A – Stated Likeliness to Vote

Respondents are asked how likely they are to vote in the upcoming election.

1. How likely would you say you are to vote in the [specify] Election?*

  • 5 – Almost certain
  • 4 – Very likely
  • 3 – Somewhat likely
  • 2 – Not very likely
  • 1 – Not at all likely

*”Already voted” is a possible response depending on when the survey is administered. In the case of respondents who say that they already voted, the respondent receives a likelihood value equivalent to the value of “almost certain”.

2. I’d like you to rate your chance of voting in the general election to be held on [specify date] on a scale of 10 to 1. If TEN represents a person who definitely will vote and ONE represents a person who definitely will NOT vote, where on this scale of 10 to 1 would you place yourself?

  • Response: (A value between 1 – 10)

3 – Calculated Likeliness to Vote

From the information gathered from the voter registration file and the information collected during the interview, the Siena College Research Institute calculates the respondent’s likelihood of voting in the upcoming election.