Overview & Tips
    Louis Harris & Associates interviewed 1,180 likely voters (aged 21+) for Newsweek magazine in November 1963 (just before President Kennedy's assassination) to explore issues including the president's performance, the US/Soviet relationship, the space program, religious matters, civil rights, the Vietnam conflict, and the Cuba situation. Overall, 57% of likely voters (40m of 71m) approved of President Kennedy's performance. We built a model to predict approval then packaged it in this simulator to estimate how changes in opinion would have affected the 57% (.57) and 40m numbers.

  • Tip 1. If you select hold constant, the simulator keeps the actual values for that predictor variable. If you select hold constant for every variable, the predicted presidential approval probability still will be .57.
  • Tip 2. If you select different values, the simulator generally assumes they apply to all likely voters, with the presidential approval probability and associated estimates changing accordingly.
  •   JFK Approval "All or Nothing" Simulator
      Check out the JFK Approval "fine-tuning" simulator.

    Harris p: 0.57  Approvers: 40m

    Applying Probabilities

    Potentially, you can apply the same predicted probability to three groups:

    1. All likely voters. If you see a way for your selected values to apply to all 71m likely voters, you can regard those values as targets to hit to increase (or decrease) the 57% (.57) and 40m numbers. To ensure they apply, select hold constant for the Respondent Race socio-demographic variable.
    2. The Target. When selecting socio-demographic criteria that do not apply to all likely voters (e.g., not everyone can be Black or White), the simulator generates a second set of results based on the actual size of the target group. For instance, if you select "Black" for Respondent Race, this second set of results will focus on the 4.2 million Black likely voters in the US in November 1963.
    3. Specific Individuals. If the values you choose are your own, or those of someone you had in mind, the resulting estimate will be your, or their, predicted probability.

    Simulation Possibilities

    You can simulate thousands of possibilities to answer questions like this:

    1. If all likely voters, rather than 9%, felt that President Kennedy did a poor job handling Soviet Union Premier Nikita Khrushchev and nothing else were to change, how would that affect the percentage (57%) and number (40m) of approvers? [The figures would decrease to 27% and 19m.]
    2. What is an individual person's presidential approval probability? [Enter values to see for yourself.]