Skip to content

Research Integrity

Research integrity is essential to everything we do in the lab. This section explains what research integrity (or, research misconduct) looks like at different stages of the research process.

General rule of thumb: It is unacceptable to falsify data, deliberately omit data or manipulate data in any way. If you are unsure about whether something is a questionable practice, please reach out to the lead researcher of the project.

Please read the following sections if you are registered for PSY 390:

  • When you are asked to test a program, make sure that you do it promptly and enter the information as instructed. When you are testing a program that asks you to enter a subject number, we will always ask you to put your name as the subject number. This is important because it allows us to easily identify and separate your data and data from actual participants .
  • Do your best to ensure that subjects are randomly assigned to conditions/versions. Before each study, we will tell you how participants should be assigned to each condition and/or version. This will depend on the design of the experiment and the tasks involved. Sometimes the design of the experiment limits how subjects can be assigned to conditions so assignment cannot be entirely random (e.g., if we have a task that requires participants to form a pair, we can run that condition when an even number of people showing up for the session). Working within those types of limitations, we try our best to make assignment random If there are systematic differences is how people are assigned to one condition compared to the other (e.g., if one experimenter always runs one condition and someone else always runs the other, or if one condition is always run in the morning and the other condition is always run in the afternoon), we wouldn’t be able to separate the effect of the experimental manipulation from the effect of the confounding variable.
  • Never tell participants about the purpose of the experiment or influence their answers in any way. Sometimes, you may know about the purpose of the experiment, or maybe even our predictions/hypotheses. That is because we would like you to understand the theoretical background of the research question. However, you should never reveal that information to the participants or guide them to answer in a specific way. If they have questions about the instructions, you should only paraphrase what is already written in the instructions. After finishing an experimental session, sometimes participants are curious about how they performed or about the research questions being investigated. You can let them know that you can’t share any additional information beyond what is in the debriefing form and remind them that it contains resources for finding more information about this type of research.
  • Ensure that data and documents are properly saved and stored. It is important for us as experimenters to know exactly how many subjects we have run for each study, and we must have a signed consent form for every single person that started the experiment (regardless of whether their data saved). We also need to know how many credits we have assigned to each person. You should notify the lead researcher immediately if one of the following situations has happened: (a)If there are subjects who showed up but did not participate in the study (whether because of an error on the experimenter’s end, or because of other situations out of our control) (b)) if there are subjects who participated in the study but their data was not saved correctly (again, whether because of an error on the experimenter’s end, or because of a technical error). You should document these instances in the end-of-session survey, as well as email the lead researchers promptly.
  • Adhere to scoring guidelines and avoid introducing bias or noise (unintended variability). Sometimes we ask you to score the answers of participants. It is important that you are consistent in how you score the responses (e.g., if you give partial credit to one person for an answer a different person should also receive partial credit for that answer). When we send you the scoring sheets, we do not disclose what condition/version they are in, and that is to ensure that your scores are not affected by that information. Even if you try to remain impartial to that information, it can influence your judgements/decisions.
  • Most importantly, if there’s an issue or mistake, contact us right away! Mistakes happen and that is understandable, but we need to know about them as soon as possible.

Please read the following section if you are registered for PSY 391 / PSY 498:

In addition to everything in the section above, you are also expected to:

  • Document the decisions we have made while designing the experiment. Some decisions about the study design need to be made before data collection (e.g., the target sample size, exclusion criteria, main dependent variable and scoring methods, etc.). While we are not entirely bound by those decisions, any modification made after data collection begins needs to be disclosed and justified.
  • Double-check and triple-check your programming code, analysis scripts and files and the statistics in your thesis. We will talk about how to insert those checks!