1.5 Contemporary Issues: A Revolution in Methodology
At the end of each chapter, the “Contemporary Issues” section addresses an issue in psychology that is going through major changes and is thought to be influential for the future. The inability to replicate a large proportion of studies in any field is highly problematic. People rely on science and the scientific method to provide accurate and reliable information, especially in the age of social media and misinformation. Increasing lack of trust in science has led people to be unwilling to wear masks during a pandemic or get themselves and their children life-saving vaccines. Therefore, there are many efforts being made to help remedy the replication crisis in psychology—let’s look at a few (refer to “The Replication Crisis: Crash Course Statistic, Crash Course ”):
Promote and value replication. There are collaborations by organizations such as the Reproducibility Project to continue focusing on and testing replication in many fields of science, including psychology. There are also efforts to emphasize replication research, including increasing funding opportunities for these experiments (Lilienfeld, 2017).Conduct meta-analyses. Recall that a meta-analysis is an evaluation of the results of many experiments that address the same or similar topics. A meta-analysis is a more reliable source for drawing conclusions and theories in psychology.
Publish null results. A null result is when the data does not support the hypothesis. Null results are typically not very flashy, nor do they make headlines, and they are generally not published. Some journals, such as the well-respected Public Library of Science (PLOS), publish research with null results if the question and methodology are sound. Having access to published papers of null results is beneficial because it allows scientists to learn from the methods used and not repeat the experiment unnecessarily.
Require submission of raw data. Many journals are now requiring that a spreadsheet of raw data be submitted along with research for review. Raw data are all the data collected from individuals before statistical analysis. The raw data are also often available when the journal is published so that any interested person can run their own statistical analysis.
Emphasize effect size. There are statistics that can measure the strength of the relationship between two variables. These statistics take into account the mean, standard deviation (variance), and sample size. For example, one antidepressant medication might be significantly better at reducing symptoms than another, but the difference might be small or large. If there is only a slight difference, one might consider other aspects of the medication, such as side effects or cost. Many journals require that the effect size of a finding be presented.
The Replication Crisis: Crash Course Statistic, Crash Course
This video is a bit longer than others, but it does an excellent job describing causes and possible remedies of the replication crisis in psychological research.
An innovative solution to the problems associated with replication and unethical data manipulation is preregistration. Suppose a psychologist or research team is starting a new project. They can preregister their research plan before collecting data. There are several preregistration websites, but one of the most often used is the Open Science Framework (OSF) website. The lead researcher can log in to OSF and describe in as much detail as they can their plans, including the main purpose of the research, the number and demographics of their participants, the material used (e.g., the survey questions), procedures for collecting data, rules for excluding participants, main hypotheses, and planned data analysis. This minimizes the researchers from distorting results and raises the credibility of the intention of the research. The information about the study is then time-stamped, saved, and made available to editors and reviewers if a manuscript is submitted to a journal. If the paper is accepted for publication, the authors and journal would then make the preregistration plan available to readers. Some journals may agree to publish the results of a preregistered experiment regardless of the outcome, which reduces the tendency to only submit positive results. The preregistration research plan can also be made available to colleagues before the research begins to elicit feedback.
Preregistration and making data available are part of the new movement of open science, where the experimental research design, purpose, data collection, and analysis are available for others to observe. The raw data from the preregistered experiment is also made open and available so that other researchers can monitor data analysis. Other researchers can also learn about novel ideas, innovative research methods, and techniques for data analysis. This open-science approach also adds commitment to ethics in research and data analysis. It can also allow others to see negative results, which may be essential in conducting reviews and meta-analyses on a topic.