Table 2.
Pre-registered research questions and hypotheses.
| question | hypothesis | sampling plan | analysis plan | interpretation given different outcomes |
|---|---|---|---|---|
| 1. do dyslexic children differ from typically developing children in motion-coherence thresholds? | dyslexic children will have higher motion-coherence thresholds than typically developing children | 48 participants per group. Power analysis based on d = 0.747, 95% power, α = 0.05 | two-tailed independent samples t-test and Bayesian t-test on motion-coherence thresholds | (a) if dyslexic children have higher motion-coherence thresholds than typically developing children (p < 0.05), this will be interpreted as reduced sensitivity in dyslexia in this task, consistent with previous research. (b) If p > 0.05 and BF10 < 1/3 (i.e. evidence in support of the null hypothesis), this suggests that the groups do not differ in motion-coherence sensitivity. (c) If p > 0.05 and BF10 > 1/3, this suggests that the data are inconclusive. (d) If dyslexic children have lower motion-coherence thresholds than typically developing children (p < 0.05), this will be interpreted as increased sensitivity in this task, in contrast with previous research |
| 2. do dyslexic children differ from typically developing children in sampling estimates in the motion-averaging task? | dyslexic children will differ from typically developing children in sampling estimates in the motion-averaging task | 48 participants per group. Power analysis based on d = 0.747, 95% power, α = 0.05 | two-tailed independent samples t-test and Bayesian t-test on sampling estimates resulting from equivalent noise analysis of performance in the motion-averaging task | (a) if dyslexic children have lower sampling estimates than typically developing children (p < 0.05), this will be interpreted as reduced motion integration. (b) If p > 0.05 and BF10 < 1/3, this suggests that the groups do not differ in motion integration abilities. (c) If p > 0.05 and BF10 > 1/3, this suggests that the data are inconclusive. (d) If dyslexic children have higher sampling estimates than typically developing children (p < 0.05), this will be interpreted as increased motion integration: a similar pattern found in autism research |
| 3. do dyslexic children differ from typically developing children in internal noise estimates in the motion-averaging task? | dyslexic children will have higher estimates of internal noise obtained from the motion-averaging task than typically developing children | 48 participants per group. Power analysis based on d = 0.747, 95% power, α = 0.05 | two-tailed independent samples t-test and Bayesian t-test on internal noise estimates resulting from equivalent noise analysis of performance in the motion-averaging task | (a) if dyslexic children have higher internal noise estimates than typically developing children (p < 0.05), this will be interpreted as reduced precision of estimating local dot directions in dyslexia. (b) If p > 0.05 and BF10 < 1/3, this suggests that the groups do not differ in terms of precision of estimating local dot directions. (c) If p > 0.05 and BF10 > 1/3, this suggests that the data are inconclusive. (d) If dyslexic children have lower internal noise estimates than typically developing children (p < 0.05), this will be interpreted as increased precision of estimating local dot directions in dyslexia |
| 4. do dyslexic children differ from typically developing children in orientation-coherence thresholds? | dyslexic children will not differ from typically developing children in orientation-coherence thresholds | 48 participants per group based on the power analysis for the motion tasks, but participants will be excluded from the orientation task analysis if they are unable to perform the orientation tasks | two-tailed independent samples t-test and Bayesian t-test on orientation-coherence thresholds | (a) if dyslexic children have higher orientation-coherence thresholds than typically developing children (p < 0.05), this will be interpreted as reduced sensitivity in dyslexia in this task. (b) If p > 0.05 and BF10 < 1/3, this suggests that the groups do not differ in orientation-coherence sensitivity. (c) If p > 0.05 and BF10 > 1/3, this suggests that the data are inconclusive. (d) If dyslexic children have lower orientation-coherence thresholds than typically developing children (p < 0.05) this will be interpreted as increased sensitivity in this task |
| 5. do dyslexic children differ from typically developing children in sampling estimates in the orientation-averaging task? | dyslexic children will not differ from typically developing children in sampling estimates in the orientation-averaging task | 48 participants per group based on the power analysis for the motion tasks, but participants will be excluded from the orientation task analysis if they are unable to perform the orientation tasks | two-tailed independent samples t-test and Bayesian t-test on sampling estimates resulting from equivalent noise analysis of performance in the orientation-averaging task | (a) if dyslexic children have lower sampling estimates than typically developing children (p < 0.05), this will be interpreted as reduced orientation integration. (b) If p > 0.05 and BF10 < 1/3, this suggests that the groups do not differ in orientation integration abilities. (c) If p > 0.05 and BF10 > 1/3, this suggests that the data are inconclusive. (d) If dyslexic children have higher sampling estimates than typically developing children (p < 0.05), this will be interpreted as increased motion integration: a similar pattern found in autism research |
| 6. do dyslexic children differ from typically developing children in internal noise estimates in the orientation-averaging task? | dyslexic children will either have higher or similar internal noise estimates compared to typically developing children | 48 participants per group based on the power analysis for the motion tasks, but participants will be excluded from the orientation task analysis if they are unable to perform the orientation tasks | two-tailed independent samples t-test and Bayesian t-test on internal noise estimates resulting from equivalent noise analysis of performance in the orientation-averaging task | (a) if dyslexic children have higher internal noise estimates than typically developing children (p < 0.05), this will be interpreted as reduced precision of estimating local orientations in dyslexia, consistent with theories of increased internal noise in dyslexia. (b) If p > 0.05 and BF10 < 1/3, this suggests that the groups do not differ in terms of precision of estimating local orientations. (c) If p > 0.05 and BF10 > 1/3, this suggests that the data are inconclusive. (d) If dyslexic children have lower internal noise estimates than typically developing children (p < 0.05), this will be interpreted as increased precision of estimating local orientations in dyslexia |