This study measured the Inference Task (created by the authors) to determine: (1) whether it measured what it was intended to measure (i.e., listening comprehension), and (2) whether questions in the Inference Task designed to assess “local” vs. “global” inferences (see right) were really different.
The authors found that the Inference Task they developed was correlated with other established tests of listening comprehension, indicating that “…the experimenter-developed Inference Task is a valid measure to assess discourse listening comprehension.”
Wait—but what is the Inference Task? Basically, short little stories with questions that follow. The best part is that the Inference Task passages, questions, and examples of answers with scores are included in the article! You could print this task out today and use it with a client. Two stories are included for each grade, preschool through 3rd (with the second story overlapping with the next grade, for 6 stories total)—meaning you could use this for progress monitoring, too. The article also includes means and standard deviations from the huge sample of children who completed the task (see article Table 1).
On whether there is a difference between local and global inferences, the short answer is maybe. The long answer is that there was some evidence that local and global inferences were distinct, but the differences were small and inconsistent. The authors propose several possible reasons for this finding; basically, local and global inferences go hand-in-hand and tap many of the same skills. So what does this mean for your clients who struggle with making inferences? The fact that different types of inference are difficult to separate means that “the training of specific functions of inferences may not affect performance on specific question types but will instead influence performance on the entire listening comprehension test.” Instead of focusing on assessing local vs. global inferences, it might be better for us to focus on improving listening comprehension overall by targeting areas of need like inference making. For more information on treatment for inference making in older children, see this recent review.
Language and Reading Research Consortium (LARRC) & Muijselaar, M. M. L. (2017). The dimensionality of inference making: Are local and global inferences distinguishable? Scientific Studies of Reading, doi: 10.1080/10888438.2017.1371179