Skip to main content


Happy almost Friday! This week, we’re covering whether sequence matters for learning outcomes. Our new article this week seeks to answer:

  • Does the order of learning tasks impact learning outcomes?

We’ll also revisit and older article, which asks:

  • Will presentation order of instructional materials impact learning?

Let’s learn about the order of learning!

Which One First? 🐔

Generative learning activities and retrieval practice both play a large role in enhancing learning outcomes (see LSW Issue #55 for more) - but what if we want to use both? Where should we put them within instructional materials? Roelle et al. (2022) aimed to evaluate if the sequence of these activities impacts learning.

After a pre-test, learners read text that covered eight concepts regarding social attribution in psychological science. Upon completion of the reading, concept-definition pairs were presented again one at a time for the learners. Learners were then split into one of three groups: “retrieval-before-generation,” “generative-before-retrieval,” or “restudy” (Roelle et al., 2022).

In this study, the retrieval practice task was purely cued recall. During the retrieval practice, concept terms were shown one at a time and learners were asked to provide the definitions “as accurately as possible.” After submitting each concept definition, ideas that should be included in the definitions were shown to learners. From there, learners submitted a feedback form where they rated whether their answer “fully, partially, or not at all” included the ideas (Roelle et al., 2022).

The generative learning task was open-book, meaning the concept definitions were shown to learners and the task did not require any retrieval. With the concept definitions shown, learners were asked to provide two of their own examples for each of the eight concepts (16 examples total). It should also be noted that a brief introductory training on how to develop quality examples was provided prior to the initial study phase (Roelle et al., 2022).

As I’m sure you likely picked up on, learners in the “retrieval-before-generation” group completed the retrieval practice task first, then the generative learning task. Learning in the “generative-before-retrieval” group completed the generative learning task first, then the retrieval task. Those in the “restudy” did not complete the retrieval or generative tasks, but were instructed to re-read the text (Roelle et al., 2022). Cognitive load was also evaluated after each of the learning tasks; see image for procedure.

Roelle et al. (2022)

The posttest timing was split between groups, with learners being randomly assigned to an immediate test (five minute delay) or a delayed test (one day delay). Learners were randomly assigned to either group, rather than taking both, due to an immediate test requiring retrieval. If learners participated in an “extra” retrieval posttest, it could impact the results of just a delayed test (Roelle et al., 2022).

So, results? To start, both experimental groups showed significantly higher gains than the control group (i.e., restudy). Ultimately, the retrieval-before-generation group outperformed the others regarding pre-post gains on cued recall, with lower cognitive load serving as a mediator for the effect (Roelle et al., 2022). Interpreting this, it is likely that the gains from pre- to post-test are due to the retrieval phase. An important note for the findings: researchers deemed the examples learners provided for the generative activity to be of fairly low quality, with only “about 41% of the idea units” being correctly illustrated (Roelle et al., 2022). Thus, understanding the role of generative activities will require more research, considering the quality of answers and their potential impact on cognitive load.

Key Takeaway(s): When implementing multiple learning strategies, learners seem to experience lower cognitive load and higher retention rates if they practice retrieval tasks before generative learning tasks.

Read More ($): Roelle, J., Froese, L., Krebs, R., Obergassel, N., & Waldeyer, J. (2022). Sequence matters! Retrieval practice before generative learning is more effective than the reverse order. Learning and Instruction, 80.

“Engaging learners in retrieval practice before engaging them in generative learning can be more beneficial than vice versa.”- Roelle et al. (2022)

⚡🔙.✌️.Issue #59

Understanding how to order practice tasks is important, but what about instructional materials? Check out Issue #59, which talks about research on the sequence of VR training.

Pets of Learning Science Weekly

Reader Elisabeth H. provided us with a perfect trio this week! The beautiful kitty is Leah, “a feisty cat who loves snuggles and being the center of attention.”

Buddy can be seen “hanging out with his friend Bear (the little doggy).” He enjoys dog parks and “hanging out with little doggies and being chased by them.” 🌳

Send us your pet pics at

Wondering why we’re including animal photos in a learning science newsletter? It may seem weird, we admit. But we’re banking on the baby schema effect and the “power of Kawaii.” So, send us your cute pet pics -- you’re helping us all learn better!

The LSW Crew

Learning Science Weekly is written and edited by Katie Vanhardt, Ph.D.

Have something to share? Want to see something in next week's issue? Send your suggestions: