People who are socially anxious are often lonely, too. But which comes first?
Loneliness is a perceived lack of social connection.
Loneliness arises when there is a mismatch between the social relationships a person desires and those they actually have. It can be temporary—such as after a move or breakup—or chronic, persisting over time and significantly affecting well-being.
Social anxiety is the state of being fearful of social situations, often out of fear of evaluation by others. Socially anxious people often avoid social situations and other situations where they might be evaluated.
Given the overlap between these two traits, a researcher might hypothesize that social anxiety leads to loneliness, in part because socially anxious people avoid being with others. But one could also argue the other direction: that loneliness leads to social anxiety, because lonely people might attribute their feelings of isolation to others’ rejection, which might build up social anxiety.
When we want to determine which psychological trait comes first in time, we can use a cross-lag panel design. That’s what a group of German researchers have done. Their study was summarized here, by the website, PsyPost.
Here’s how the journalist describes the study. I’ve removed some distracting details.
Study author Anna C. Reinwarth and her colleagues set out to explore whether and how loneliness affects symptoms of social anxiety—and vice versa—over a five-year period in a large sample drawn from the German population. […]
The researchers analyzed data from the ongoing Gutenberg Health Study, a large-scale longitudinal project based in the Rhine-Main region of Germany. […] Participants were randomly selected from local registries in Mainz and the surrounding district of Mainz-Bingen, with sampling stratified by age and sex to ensure representativeness. […]
Participants provided sociodemographic information and completed assessments measuring loneliness (via a single-item question: “I am frequently alone/have few contacts”), [and] social anxiety symptoms (using the three-item short form of the Social Phobia Inventory), […]
Data for the current analysis came from two assessment waves: the baseline wave conducted between 2007 and 2012, and a follow-up wave conducted five years later between 2012 and 2017. In total, 15,010 people participated in the baseline wave, and 12,423 were followed up five years later. […]
The results showed that 11% of participants reported feelings of loneliness, and 7% reported symptoms indicative of social anxiety. Over the five-year period, symptoms of social anxiety were found to be more stable than loneliness, although both showed some consistency over time.
Participants who reported higher levels of social anxiety symptoms at baseline were more likely to report increased loneliness five years later. However, initial feelings of loneliness were not significantly associated with later increases in social anxiety symptoms after adjusting for relevant factors.
Questions:
a) Classify the two main variables in the study using the following table:
| Variable name (Stated at the conceptual level) | How was this variable operationalized? | Potential levels of this variable (if you don’t know the numbers, you can say “low to high” | Is this variable manipulated or measured? |
b) Was this study correlational or experimental? Explain your answer.
c) In the texbook (Chapter 9) the example of a cross-lag panel design had collected data at four time points. How many time points were used in the present study?
d) Now, following the figures in the textbook (e.g., Figure 9.3), sketch and label the boxes of the cross-lag panel design of this study. (Remember, this one doesn’t have four time points.)
e) Next, what is the summary saying about the autocorrelations? Where will these autocorrelations go on your sketch? Add some arrows to your sketch, and make up some correlations (r)s to estimate what the autocorrelations might be–use Figure 9.2 as your model. Read the summary to find out which autocorrelation is higher.
f) Next, sketch two cross-lag arrows on your diagram, leaving a spot for each cross-lag correlation (use Figure 9.3 as your model). Which of these cross-lag correlations was higher, according to the summary? Make up some correlation (r)s to estimate what the cross-lag correlations might be.
g) Using the original empirical article, compare your sketch to the study’s actual results. Scroll to the article’s Figure 1. The relationships found in the study are positioned on top of each of the four arrows. You should look at the values labeled M1 in the top row (these are the unstandarized coefficients, without any covariates, so they are comparable to the rs you estimated).
h) Reflect on the pattern. Explain, in your own words, why the pattern of cross-lag correlations is consistent with the conclusion in the PsyPost headline: “Social anxiety predicts future loneliness, study finds — but not the other way around”.
i) Bonus question: Locate the information in the summary that can tell you about the external validity of this study.