Causal and Non-Causal Language in reporting regressions

When reporting the results of a regression analysis, it is important to be careful about the language you use. Regression is a statistical technique that measures the relationship between one or more predictor variables and an outcome variable. However, this relationship does not necessarily imply causation, meaning that changes in one variable cause changes in another. There may be other factors that influence both variables, or the direction of causality may be reversed, or the relationship may be spurious or coincidental.

  1. Spurious Relationship: This occurs when two variables appear to be related to each other but are actually both influenced by a third variable, which is the true cause of the observed relationship. The classic example is the relationship between ice cream sales and drowning incidents. There is a positive correlation between these two variables – as ice cream sales increase, so do drowning incidents. However, this does not mean that buying ice cream causes drowning. The third variable, which is the actual cause of both, is the temperature or season (specifically, summer). During summer, more people buy ice cream, and more people also go swimming, which unfortunately increases the risk of drowning.
  2. Reversed Relationship: This refers to a situation where the direction of causality between two variables is opposite to what is commonly believed or hypothesized. A classic example is the relationship between self-esteem and success. It is often believed that high self-esteem leads to success in life. However, some research suggests that this relationship might be reversed – it could be that success in various endeavors leads to high self-esteem, not the other way around. In other words, the causal direction is from success to self-esteem, rather than from self-esteem to success.

Therefore, when writing about regression results, you are recommended to avoid using causal language unless you have strong evidence to support a causal claim. Causal language includes words or phrases that suggest a cause-and-effect relationship, such as “effect”, “impact”, “influence”, “determine”, “cause”, “lead to”, “result in”, etc. Instead, you will be required to use non-causal language that describes the association or correlation between variables, such as “associated with”, “related to”, “linked to”, “predict”, “explain”, “account for”, etc.

Using non-causal language does not mean that you are denying the possibility of causation, but rather that you are acknowledging the limitations of your study design and data. By being precise and cautious in your language, you can avoid making unwarranted or misleading claims that may confuse or misinform your readers. In this blog post, we will review some examples of causal and non-causal language in reporting regression results, and provide some tips on how to choose the appropriate language for your research context.

Please find the following post with the flowchart that you can decide whether you can use the causal words or not.

Sargeant et al. (2022) provides the list of non-causal and causal words as follows.

Non-causal words

  • Association/associated
  • Risk factor/sparing factor/protective factor
  • Higher/lower/more or less likely/greater/lesser
  • Predicts/predictor/prognostic factor
  • Related to/relation(ship)
  • Correlated (with)
  • Explain(ed)/explanatory (variables)
  • Linked to
  • Varies with

Causal words

  • Increases/decreases
  • Effects/affects/is effective
  • Influences
  • Elevates/reduces/depresses
  • Impacts
  • Contributes to
  • Cause(d)/consequence of
  • Is attributed to
  • Leads to
  • Improves/results in improvement
  • Is responsible for
  • Prevents
  • Results in
  • Proved to be
  • Altered by
  • Induces (an outcome)
  • Is successful
  • Enhances
  • Benefits from
  • Impairs
  • Driver of
  • Underlies

Qualifying words

  • Can/could
  • May/might
  • Suggests that
  • Likely that
  • Appears to be
  • Possible that
  • Seems to
  • Presents some evidence
  • Thought to be


Sargeant, J. M., O’Connor, A. M., Totton, S. C., & Vriezen, E. R. (2022). Watch your language: An exploration of the use of causal wording in veterinary observational research. Frontiers in Veterinary Science9, 1004801.

Thapa, D. K., Visentin, D. C., Hunt, G. E., Watson, R., & Cleary, M. (2020). Being honest with causal language in writing for publication. Journal of advanced nursing76(6), 1285-1288.

  • January 18, 2024