Retrieval Practice with First Graders (6-7 years)

Retrieval Practice with First Graders (6-7 years)

In today’s post, I present a paper from 2020 by Xiaofeng Ma and colleagues (1) investigating retrieval practice with first grade students (children ages 6-7 years). They wondered whether retrieval practice would be effective for children this young, given that children’s memory systems are less mature at this age. Children this age are much less likely to be able to engage in free recall or answer multiple-choice and short-answer questions in the same way as older children or adult learners. So, they modified the retrieval activity so that it was developmentally appropriate for 6 and 7 year old students. (For another example of modifying retrieval practice for kids, see this post as well.)

The Experiment Method

In Ma and colleagues’ experiment (1), first grade students in an elementary school in China learned the objects in 15 different pictures. The pictures involved one of three backgrounds (a kitchen, garden, or bedroom). Each picture depicted one object within the background that one might find in those spaces (e.g., a bowl, a butterfly, slippers). There were five different pictures of objects using each background, creating a total of 15 different pictures.

The students learned the pictures in one of four different learning conditions. Time was equated, such that each student spent 20 seconds (or an average of 20 seconds) learning each picture.

Retrieval practice with feedback: The students saw a picture for 5 seconds. Then the picture was removed, and they had to retrieve what the picture was (4 seconds). Then, as feedback, they saw the picture again briefly (1 second). They repeated this process one more time with the same picture (saw the picture, retrieved the object, saw the picture again as feedback, 10 seconds total). They learned each of the 15 pictures in this way.

Retrieval practice without feedback: The students saw a picture for 5 seconds, and then were given three opportunities to try to recall the picture (5 seconds each). No feedback was given in this condition, and they learned each of the 15 pictures in this way.

Elaboration: The students were told they were playing a game on the computer. To learn the objects that went in each background, they were allowed to click and drag the objects, moving them around to create their own short story. They spent 100 seconds total with each background (an average of 20 seconds for each of the 5 objects). They played this game with all three backgrounds.

Repeated viewing: The students saw each picture four times in a row, each time for 5 seconds. They learned each of the 15 pictures in this way.

The researchers also manipulated retention interval (i.e., the delay, or amount of time between learning and measuring retention). After they learned the objects from the pictures using their assigned learning condition, students’ memory for the objects was tested 5 minutes later, 1 week later, and 1 month later. During the test, students saw 15 pictures that they learned and 15 new pictures they did not study. Their job was to tell the researcher which ones they had learned before.


In short, retrieval practice with feedback led to the best memory accuracy overall.

When a student correctly identifies one of the pictures that they studied, this is called a “hit” in cognitive psychology. When a student incorrectly identifies a new picture as one that was studied, this is called a “false alarm” in cognitive psychology. Ma and colleagues analyzed the hits and false alarms separately.

Numerically, retrieval practice with feedback resulted in the highest recognition of the pictures that were studied (hits) and the lowest incorrect identification of new pictures as studied (false alarms) for all three retention intervals.

There were a number of statistically significant results, but there are a few notable highlights related to retrieval practice. Retrieval practice with feedback led to significantly higher hit rates (so, more correct identifications) than the other three learning conditions across all three of the retention intervals. Retrieval practice without feedback led to greater hit rates than repeated viewing after 1 month.