PSYC 255 Week 4 DB Forum 3 Example

Discussion Board Forum 3: Thread and Two Replies

PSYC 255 Introduction to Research, Discussion Board Forum 3 (thread and two replies)

Thread by [Student Name]

The prompt, the example and the classmates' posts are composites standing in for the course's own forum; the textbooks, Scripture and published article are real; the requirements follow Liberty's public course guide for PSYC 255.

Thread

Last week a news site in my feed announced that phones and apps had set off an epidemic of teenage depression. Many studies do find that teenagers who spend more time on social media report more symptoms of depression, and the pattern is worrying enough to deserve attention. But the headline makes a causal claim, and a correlation alone cannot support it.

Crawford (2019) explains that a correlation tells us that two variables move together, not why. At least three other explanations remain open. The first is directionality: depressed teenagers may withdraw from in-person activities and spend more time scrolling, so depression could increase social media use rather than the reverse. The second is a third variable: poor sleep, for example, might both increase late-night phone use and worsen mood. The third is selection: if a study recruits only teens who already visit a clinic, the sample may create a link that does not exist in the general population.

Researchers often try to rule out third variables by statistically controlling for them, but that strategy has its own risks. Rohrer (2018) explains that graphical causal models, diagrams with arrows showing how variables might influence one another, help researchers decide what to control. Controlling for a genuine confound improves an estimate, but controlling for a mediator, a variable on the causal path, or a collider, a variable caused by both, can move the estimate further from the truth. Causal conclusions from observational data always depend on assumptions that must be stated. In the social media example, controlling for loneliness could be a mistake if social media use causes loneliness, which in turn worsens mood; the control would hide part of the very effect being studied.

In Proverbs 18:17, an argument that sounds airtight on its own falls apart once the other party gets to cross-examine it (English Standard Version Bible, 2001/2016). Headlines are the opening argument; research methods are the cross-examination.

Classmates: what other headline would you diagram this way, and which arrow do you think is most often missed?

Reply to Natalie

Natalie, you asked how researchers could ever know whether social media causes depression, given all the alternatives. The strongest answer is an experiment with random assignment. If researchers randomly assigned a group of teenagers to reduce their social media use for several weeks and compared them with a group who continued as usual, random assignment would spread differences in sleep, family stress and prior mood evenly across the groups. Any difference in depression afterward could then be attributed more confidently to the change in use. Crawford (2019) describes random assignment as the feature that allows experiments to support causal conclusions. Such experiments have limits, since short reductions may not reflect years of use, but they answer a different question than surveys can. Would you trust a two-week experiment to tell us about long-term effects, or would you want to see longitudinal data as well?

Reply to Isaiah

Isaiah, I appreciated your point that research methods teach intellectual humility, because they make us admit what a study cannot show. Rohrer (2018) makes a similar argument: drawing causal conclusions from observational data is never a mechanical procedure but always depends on assumptions that require knowledge of the topic, and the same is true, in different ways, of all empirical research. The humble response is not to stop drawing conclusions but to state our assumptions clearly so that others can test them, and to change our minds when better evidence arrives. The instruction in James 1:19 to listen readily and talk reluctantly (English Standard Version Bible, 2001/2016) might be good advice for anyone writing a headline about a new study. How do you think researchers could communicate uncertainty to the public without sounding as if they know nothing?

References

Crawford, R. (2019). Introduction to research: Less fright, more insight. Kendall Hunt.

English Standard Version Bible. (2016). Crossway. (Original work published 2001)

Rohrer, J. M. (2018). Thinking clearly about correlations and causation: Graphical causal models for observational data. Advances in Methods and Practices in Psychological Science, 1(1), 27-42. https://doi.org/10.1177/2515245917745629