analyses:viechtbauer2007b
Differences
This shows you the differences between two versions of the page.
Both sides previous revisionPrevious revisionNext revision | Previous revisionLast revisionBoth sides next revision | ||
analyses:viechtbauer2007b [2020/06/26 06:54] – Wolfgang Viechtbauer | analyses:viechtbauer2007b [2022/03/26 15:28] – Wolfgang Viechtbauer | ||
---|---|---|---|
Line 11: | Line 11: | ||
dat <- dat[c(7: | dat <- dat[c(7: | ||
dat$dosage <- (dat$dosage * 7) / 1000 | dat$dosage <- (dat$dosage * 7) / 1000 | ||
+ | rownames(dat) <- 1:nrow(dat) | ||
dat | dat | ||
</ | </ | ||
Line 34: | Line 35: | ||
17 98 186 80 189 0.2189 0.0120 | 17 98 186 80 189 0.2189 0.0120 | ||
</ | </ | ||
- | Variables '' | + | Variables '' |
Note that, for illustration purposes, only a subset of the data from the Linde et al. (2005) meta-analysis are actually included in this example. Therefore, no substantive interpretations should be attached to the results of the analyses given below. | Note that, for illustration purposes, only a subset of the data from the Linde et al. (2005) meta-analysis are actually included in this example. Therefore, no substantive interpretations should be attached to the results of the analyses given below. | ||
Line 45: | Line 46: | ||
</ | </ | ||
<code output> | <code output> | ||
- | ai n1i ci n2i | + | ai n1i ci n2i yi dosage major baseline duration ci.lb ci.ub |
- | 1 20 25 11 25 1.82 0.06 2.66 | + | 1 20 25 11 25 1.82 |
- | 2 14 20 9 20 1.56 0.08 6.30 | + | 2 14 20 9 20 1.56 |
. | . | ||
- | 13 55 123 57 124 0.97 0.02 6.30 | + | 13 55 123 57 124 0.97 |
. | . | ||
- | 17 98 186 80 189 1.24 0.01 6.30 | + | 17 98 186 80 189 1.24 |
</ | </ | ||
With '' | With '' | ||
Line 59: | Line 60: | ||
The first model discussed in the article assumes that the //true// relative improvement rates are identical in the various studies and the only reason why the // | The first model discussed in the article assumes that the //true// relative improvement rates are identical in the various studies and the only reason why the // | ||
<code rsplus> | <code rsplus> | ||
- | res <- rma(yi, vi, data=dat, method=" | + | res <- rma(yi, vi, data=dat, method=" |
res | res | ||
</ | </ | ||
<code output> | <code output> | ||
- | Fixed-Effects Model (k = 17) | + | Equal-Effects Model (k = 17) |
- | Test for Heterogeneity: | + | I^2 (total heterogeneity / total variability): |
+ | H^2 (total variability / sampling variability): | ||
+ | |||
+ | Test for Heterogeneity: | ||
Q(df = 16) = 51.55, p-val < .01 | Q(df = 16) = 51.55, p-val < .01 | ||
Model Results: | Model Results: | ||
- | estimate | + | estimate |
- | 0.33 | + | 0.33 0.05 6.78 < |
--- | --- | ||
- | Signif. codes: | + | Signif. codes: |
</ | </ | ||
Since we are analyzing the log of the relative improvement rates, the model estimate also reflects the log relative rate. For easier interpretation, | Since we are analyzing the log of the relative improvement rates, the model estimate also reflects the log relative rate. For easier interpretation, | ||
Line 90: | Line 94: | ||
Given that the true (log) relative rates are apparently heterogeneous, | Given that the true (log) relative rates are apparently heterogeneous, | ||
- | - We can interpret the model estimate obtained above as an estimate of the (weighted) average of the true log relative rates for these 17 studies. This is the so-called fixed-effects model, which allows us to make a // | + | - We can interpret the model estimate obtained above as an estimate of the (weighted) average of the true log relative rates for these 17 studies. This is the so-called fixed-effects model, which allows us to make a // |
- We can model the heterogeneity in the true log relative rates and apply a random-effects model. This allows us to make an // | - We can model the heterogeneity in the true log relative rates and apply a random-effects model. This allows us to make an // | ||
Line 107: | Line 111: | ||
H^2 (total variability / sampling variability): | H^2 (total variability / sampling variability): | ||
- | Test for Heterogeneity: | + | Test for Heterogeneity: |
Q(df = 16) = 51.55, p-val < .01 | Q(df = 16) = 51.55, p-val < .01 | ||
Model Results: | Model Results: | ||
- | estimate | + | estimate |
- | 0.45 | + | 0.45 0.09 4.87 < |
--- | --- | ||
Line 134: | Line 138: | ||
dat$dosage <- dat$dosage * dat$duration | dat$dosage <- dat$dosage * dat$duration | ||
</ | </ | ||
- | The baseline HRSD score will be used to reflect the severity of the depression in the patients. Since these two variables may interact, their product will also be included in the model. Finally, for easier interpretation, | + | The baseline HRSD score will be used to reflect the severity of the depression in the patients. Since these two variables may interact, their product will also be included in the model. Finally, for easier interpretation, |
We can fit a mixed-effects meta-regression model with these moderators to the data with: | We can fit a mixed-effects meta-regression model with these moderators to the data with: | ||
Line 151: | Line 155: | ||
R^2 (amount of heterogeneity accounted for): 47.38% | R^2 (amount of heterogeneity accounted for): 47.38% | ||
- | Test for Residual Heterogeneity: | + | Test for Residual Heterogeneity: |
QE(df = 13) = 27.9903, p-val = 0.0091 | QE(df = 13) = 27.9903, p-val = 0.0091 | ||
- | Test of Moderators (coefficient(s) | + | Test of Moderators (coefficients |
QM(df = 3) = 10.1280, p-val = 0.0175 | QM(df = 3) = 10.1280, p-val = 0.0175 | ||
Model Results: | Model Results: | ||
- | | + | |
- | intrcpt | + | intrcpt |
- | I(dosage - 34) -0.0058 | + | I(dosage - 34) -0.0058 |
- | I(baseline - 20) -0.0672 | + | I(baseline - 20) -0.0672 |
- | I(dosage - 34): | + | I(dosage - 34): |
--- | --- | ||
Line 183: | Line 187: | ||
| | ||
</ | </ | ||
- | So, for a low baseline HRSD score (i.e., mildly depressed patients), the estimated average relative improvement rate is quite high (2.67 with 95% CI: 1.46 to 4.88), but at a high baseline HRSD score (i.e., more severely depressed patients), the estimated average relative improvement rate is low (1.26 with 95% CI: 0.99 to 1.61) and in fact not significantly different from 1. | + | So, for a low baseline HRSD score (i.e., |
As shown in Figure 3 in the article, we can illustrate these results with a scatterplot of the data, superimposing a line (or rather: curve after the back-transformation) with the estimated average relative improvement rate based on the model for different baseline HRSD scores, holding the total dosage value constant at 34. This figure can be re-created with: | As shown in Figure 3 in the article, we can illustrate these results with a scatterplot of the data, superimposing a line (or rather: curve after the back-transformation) with the estimated average relative improvement rate based on the model for different baseline HRSD scores, holding the total dosage value constant at 34. This figure can be re-created with: | ||
<code rsplus> | <code rsplus> | ||
- | size <- 1 / sqrt(dat$vi) | + | xvals <- seq(12, 24, by=0.1) - 20 |
- | size <- 0.15 * size / max(size) | + | modvals <- cbind(0, cbind(xvals, 0)) |
- | + | preds <- predict(res, | |
- | modvals <- cbind(0, cbind(seq(12, 24, by=0.1)) - 20, 0) | + | |
- | preds <- predict(res, | + | |
- | plot(NA, NA, xlab=" | + | regplot(res, mod=3, pred=preds, xvals=xvals, |
- | abline(h=seq(1, 4, by=0.5), col="lightgray") | + | shade=FALSE, |
- | abline(v=seq(14, 24, by=2), col=" | + | |
- | lines(modvals[, | + | xlab=" |
- | lines(modvals[,2] + 20, preds$ci.lb, | + | axis(side=1, at=seq(12, 24, by=2) - 20, labels=seq(12, 24, by=2)) |
- | lines(modvals[, | + | |
- | symbols(dat$baseline, | + | |
</ | </ | ||
analyses/viechtbauer2007b.txt · Last modified: 2022/08/03 11:24 by Wolfgang Viechtbauer