Informed consent is the principle that states that—to the best of a psychologists' ability—plain language that is understandable to the subject is to be used to gain the consent of persons upon whom research, treatment, or services are to be performed. In instances where a person is legally incapable of providing consent, psychologists are still required to provide an appropriately phrased explanation, obtain the assent of the person involved, consider that person's preferences, and seek authorization from any guardian or responsible third party. An unconscious person may give consent prior to becoming unconscious (e.g. in sleep research and certain forms of therapy), or have a responsible guardian give assent for them. Although many ethical principles for animal care apply, a chimpanzee is a non-person and the research associated with these organisms does not require informed consent. A four-year-old child can still have most procedures explained to them in simplified language and they can voice their preference for procedures—with proper parental/guardian consent likely also required. Even an unwilling prisoner may have certain court-ordered procedures performed on them (e.g. in the mandatory psychological evaluation provided for potential suicide/self-harm cases). In these cases, every effort must be made to ensure the person understands what is going to happen before it takes place and has the opportunity to voice their concerns or questions.

The principle of confidentiality covers situations in which the law requires certain forms of disclosure. This includes limited circumstances that force the researcher to provide personally identifiable information in order to prevent harm to persons or to obtain consultations for therapy/medical treatments.

When a study is published, a researcher may rephrase or simplify the language, results, conclusions, or methods of the research in order to appeal to a wider audience. Sometimes, particular journals require the author to simplify research before a submission can be accepted. This is only acceptable when no relevant facts are changed and the rephrasing of conclusions are not suggestive.

The principle of confidentiality requires all psychologists to take steps before, during, and after research to ensure that no individually identifiable information is accessible to non-researchers without the expressed consent of the participants. A case study can be defined as an intensive review of at least one person’s psychological profile. It is virtually impossible to remove all identifiable information from such a report. For this reason, confidentiality agreements are often lengthy in case studies, and the participant(s) must be kept aware of the risk that, in spite of a researcher's best efforts, their identity may become known.

The junior researcher is forgetting all of these concepts of psychological research; furthermore, post-deception the experiment must be followed with a debriefing, in which participants are alerted to the true nature of their participation.

Nearly all research in psychology must be approved ahead of time by either an IRB or ethics board, in order to determine what risks are present in the experimental design and whether proper procedures and practices are being implemented. Examples of research that might not require initial approval are archival research, meta-analyses (though these often eventually require IRB approval for publication), and research using only computer simulations as participants.

In addition to the oversight of the humane treatment of animals during experimentation, each university or research institute's IACUC is also responsible for approving research to be performed on animals, reporting on any preventable animal deaths, and serving as a liaison for animal ethics issues between governing bodies like the APA and colleges.

Experimenter bias is the process by which an informed researcher unconsciously shows preference for one or more groups, based on knowledge outside the experiment. Note that this is not a choice—deliberately favoring one group due to outside knowledge is called fraud, and is definitely an ethics violation. A double blind methodology could rectify this issue.

The principle of informed consent requires that deception only occur when there is no reasonable non-deceptive alternative. Cost may be a factor in making this determination; however, it is rare for deception to be approved due to cost factors alone. This may also be considered when the research has a significantly established or reasonably expected value: academic or medical.

The principle of protection from harm possesses four requirements. First, subjects should be safeguarded from physical and mental harm at all times. Second, debriefing after an experiment should include a therapeutic safeguard if necessary. Third, if danger is present in an experiment, then an ethics board must approve it. Last, any danger present in an experiment must be kept at the minimum required for the experiment to function.