Table 5. Quality of evidence (GRADE) rating table.
For each outcome, the initial rating is high. This was increased or decreased according to the ratings of six factors (see following notes). The final rating is high, medium, or low quality of evidence, which defines the certainty of each outcome, which is based on the guidelines of Ryan, Santesso & Hill (2016). The following criteria were used to calculate the ratings (McArthur et al., 2018, Table 6): “Note. 1. Risk of bias: No downgrade (0) if 75% + studies contributing to an outcome are low in majority of biases. Downgrade one level (−1) if 50% to 74% of studies contributing to an outcome are low in majority of biases. Downgrade two levels (−2) if fewer than 50% studies contributing to an outcome are low in majority of biases. 2. Heterogeneity: No downgrade (0) if I2 less than 70% or I2 greater than 70% but assessment of heterogeneity and sensitivity analyses suggest the outcome is reliable. Downgraded one level (−1) if I2 70% to 85% and heterogeneity and sensitivity analyses suggest that it does affect reliability of results. Downgraded two levels (−2) if I2 greater than 85% and heterogeneity and sensitivity analyses suggest it does affect reliability of results. 3. Indirectness: No downgrade if study directly measures outcomes of interest in the population of interest. Downgraded by one level if outcome or population are not measured directly. Downgraded two levels (−2) if outcome and population are not measured directly. 4. Imprecision: No downgrade (0) if confidence interval 0 to 0.3. Downgrade one level (−1) if confidence interval 0.3 to 0.6. Downgrade two levels (−2) if confidence interval 0.6 +. 5. Publication bias: No downgrade (0) if funnel plot done on more than 10 studies (Sterne, Egger & Moher, 2011), and no bias detected. Downgrade one level (−1) if funnel plot cannot be constructed (too few studies) but bias not suspected. Downgrade two levels (−2) if funnel plot not possible (too few studies) and bias suspected. 6. Other factors: Upgrade one level (+1) if large effect size (0.8+) or no plausible confounds.
| Self-concept Outcome | Risk of Bias1 | Heterogeneity2 | Indirectness3 | Imprecision4 | Publication bias5 |
Other6 | GRADE | Certainty of outcome |
|---|---|---|---|---|---|---|---|---|
| Primary outcome | ||||||||
| Average | 0 | 0 | 0 | −1 | 0 | 0 | Medium | Probable |
| Secondary outcomes | ||||||||
| Reading-Spelling-Writing | 0 | 0 | 0 | −2 | −1 | +1 | Low | Possible |
| Academic | 0 | 0 | 0 | −2 | −1 | 0 | Low | Possible |
| Math | 0 | 0 | 0 | −2 | −1 | 0 | Low | Possible |
| Global | 0 | 0 | 0 | −1 | −1 | 0 | Low | Possible |
| Behavioural | 0 | 0 | 0 | −1 | −1 | 0 | Low | Possible |
| School | 0 | 0 | 0 | −1 | −1 | 0 | Low | Possible |
| Physical Appearance | 0 | 0 | 0 | −1 | −1 | 0 | Low | Possible |
| Work | 0 | 0 | 0 | −2 | −1 | 0 | Low | Possible |
| Social | 0 | 0 | 0 | −1 | −1 | 0 | Low | Possible |
| Athletic | 0 | 0 | 0 | −1 | −1 | 0 | Low | Possible |
| Home | 0 | 0 | 0 | −2 | −1 | 0 | Low | Possible |