In response to the question, “Have you ever committed shoplifting,” not everybody will always give an honest answer! Randomized response is a questioning technique which has been designed to obtain reliable responses on questions about a sensitive topic.
With Randomized Response there are two types of yes-no questions which are asked: the real question of interest and an alternative question. The real or the alternative question is allocated to the interviewee by some randomized technique, such as choosing a green or red card from a box of cards - the interviewer does not know which question the interviewee is answering, and so the answer cannot be used against the interviewee. The answer on the real question should give an indication of the yes-no distribution on the (sensitive) topic. The distribution of the answers to the alternative question (for example: throw with a die, did a 2 come up?) is known. If the percentage of alternative questions is known (for example, from the numbers of green and red cards in the box) then it is possible to reconstruct from the results an estimate of the distribution of the real question.
The simulation shows how this technique works for different values of the parameters. The parameters are:
| - | The number of subjects |
| - | The probability of “yes” with the real question; in practice this is the value you are looking for. |
| - | The percentage of people answering the alternative question |
| - | The probability of yes with the alternative question. |
In the table, the Expected values are the theoretical values.
The results of the simulation can be recorded in several ways:
| - | in a summary table, with the estimate based on the experiment. |
| - | in a (stacked) bar chart. The dark part refers to the yes-answers to the real question |
| - | in Statistics from all experiments the results are from the estimates obtained from all the experiments. |
Settings and buttons
Settings
In the settings window you choose the values of the parameters.
Tempo bar
With the tempo bar you can change the speed of the simulation before and during the simulation.
With Step you can track the experiment step by step.
Fast is designed to enable you to create lots of data.
Interrupt/continue
With the space bar you can interrupt or resume the simulation.
saves the results in a file.
This file can be opened and edited in the module Data analysis.