| Condition | M_Harm | M_Lie | M_Censor | Behave |
|---|---|---|---|---|
| Fact | 3.33 | 4.00 | 5.00 | 6 |
| Fact | 6.00 | 6.67 | 5.50 | 7 |
| Fact | 3.00 | 3.33 | 2.25 | 3 |
| Experience | 1.00 | 1.00 | 1.00 | 1 |
| Experience | 7.00 | 4.00 | 1.00 | 1 |
| Experience | 3.67 | 4.00 | 2.00 | 3 |
Methoden der empirischen Kommunikations- und Medienforschung
Freie Universität Berlin
| Condition | M_Harm | M_Lie | M_Censor | Behave |
|---|---|---|---|---|
| Fact | 3.33 | 4.00 | 5.00 | 6 |
| Fact | 6.00 | 6.67 | 5.50 | 7 |
| Fact | 3.00 | 3.33 | 2.25 | 3 |
| Experience | 1.00 | 1.00 | 1.00 | 1 |
| Experience | 7.00 | 4.00 | 1.00 | 1 |
| Experience | 3.67 | 4.00 | 2.00 | 3 |
Morgan wants to widely share her [pro-life | pro-choice] views. To what extent would you …
Censor_1)Censor_2)Censor_3)Censor_4)1–7, very unwilling to very willing.
| Censor_1 | Censor_2 | Censor_3 | Censor_4 |
|---|---|---|---|
| 2 | 2 | 2 | 2 |
| 2 | 2 | 2 | 2 |
| 2 | 2 | 2 | 2 |
| 4 | 4 | 4 | 4 |
| 2 | 2 | 2 | 2 |
| 1 | 1 | 1 | 1 |
| 3 | 3 | 3 | 3 |
| 2 | 2 | 3 | 4 |
| 4 | 3 | 3 | 3 |
| 5 | 2 | 2 | 2 |
| Parameter | Censor_4 | Censor_3 | Censor_2 |
|---|---|---|---|
| Censor_1 | 0.87*** | 0.88*** | 0.91*** |
| Censor_2 | 0.92*** | 0.93*** | |
| Censor_3 | 0.93*** |
p-value adjustment method: Holm (1979)
Cronbachs \(\alpha\) [1] 0.9743875
| Censor_1 | Censor_2 | Censor_3 | Censor_4 | M_Censor |
|---|---|---|---|---|
| 2 | 2 | 2 | 2 | 2.00 |
| 4 | 3 | 3 | 2 | 3.00 |
| 2 | 2 | 2 | 2 | 2.00 |
| 2 | 2 | 2 | 2 | 2.00 |
| 3 | 5 | 6 | 5 | 4.75 |
| 1 | 1 | 1 | 1 | 1.00 |
| 5 | 5 | 3 | 4 | 4.25 |
| 5 | 4 | 4 | 5 | 4.50 |
| 2 | 2 | 2 | 2 | 2.00 |
| 2 | 2 | 2 | 2 | 2.00 |
Sanjay Srivastava on Twitter (now deleted)

{lavaan}
| lhs | op | rhs | est | se | z | pvalue | ci.lower | ci.upper | std.all |
|---|---|---|---|---|---|---|---|---|---|
| Censor | =~ | Censor_1 | 1.53 | 0.06 | 25.14 | 0 | 1.41 | 1.65 | 0.92 |
| Censor | =~ | Censor_2 | 1.55 | 0.06 | 24.74 | 0 | 1.43 | 1.67 | 0.97 |
| Censor | =~ | Censor_3 | 1.58 | 0.06 | 25.01 | 0 | 1.45 | 1.70 | 0.96 |
| Censor | =~ | Censor_4 | 1.62 | 0.06 | 27.10 | 0 | 1.50 | 1.73 | 0.95 |
std.all quantifizieren den Einfluss des latenten Konstrukts auf die Items. Sie liegen in der Regel zwischen \(-1\) und \(+1\).| lhs | op | rhs | est | se | z | pvalue | ci.lower | ci.upper | std.all |
|---|---|---|---|---|---|---|---|---|---|
| Censor_1 | Censor_1 | 0.41 | 0.09 | 4.72 | 0.00 | 0.24 | 0.57 | 0.15 | |
| Censor_2 | Censor_2 | 0.17 | 0.04 | 4.02 | 0.00 | 0.09 | 0.25 | 0.07 | |
| Censor_3 | Censor_3 | 0.19 | 0.04 | 4.70 | 0.00 | 0.11 | 0.27 | 0.07 | |
| Censor_4 | Censor_4 | 0.26 | 0.06 | 4.39 | 0.00 | 0.14 | 0.37 | 0.09 | |
| Censor | Censor | 1.00 | 0.00 | 1.00 | 1.00 | 1.00 |
std.all den nicht durch die latente Variable erklärten Varianzanteil. 1 minus dieser Wert = \(R^2\). Sie liegen in der Regel zwischen \(0\) und \(1\).$Censor
Composite `Censor` is composed of observed variables:
Censor_1, Censor_2, Censor_3, Censor_4
True-score variance is represented by common factor(s):
Censor
Total variance of composite `Censor` determined from the unrestricted model.
The proportion attributable to "true" scores is its model-based estimate of reliability ("omega"):
[1] 0.975
chisq.scaled 4.176
df.scaled 2.000
pvalue.scaled 0.124
rmsea.robust 0.102
rmsea.ci.lower.robust 0.000
rmsea.ci.upper.robust 0.241
srmr 0.006
cfi.robust 0.996
{dynamic} und Web App.Your DFI cutoffs:
SRMR RMSEA CFI
Level-0 0.007 0.088 0.997
Specificity 95% 95% 95%
Level-1 0.007 0.089 0.997
Sensitivity 65% 72% 72%
Empirical fit indices:
Chi-Square df p-value SRMR RMSEA CFI
4.176 2 0.124 0.006 0.102 0.996
Notes:
-'Sensitivity' is % of hypothetically misspecified models correctly identified by cutoff in DFI simulation
-Cutoffs with 95% sensitivity are reported when possible
-If sensitivity is <50%, cutoffs will be supressed
| lhs | op | rhs | mi | epc | sepc.lv | sepc.all | sepc.nox |
|---|---|---|---|---|---|---|---|
| Censor_3 | Censor_4 | 16.37 | 0.09 | 0.09 | 0.41 | 0.41 | |
| Censor_1 | Censor_2 | 16.37 | 0.08 | 0.08 | 0.32 | 0.32 | |
| Censor_1 | Censor_4 | 6.02 | -0.05 | -0.05 | -0.17 | -0.17 | |
| Censor_2 | Censor_3 | 6.02 | -0.05 | -0.05 | -0.30 | -0.30 | |
| Censor_2 | Censor_4 | 2.83 | -0.04 | -0.04 | -0.18 | -0.18 | |
| Censor_1 | Censor_3 | 2.83 | -0.04 | -0.04 | -0.13 | -0.13 |
mi sind Differenzen in \(\chi^2\) (Betrag inhaltlich nicht interpretierbar; ab etwa 4 relevant).{lavaan}Kongenerisches Messmodell
\(\tau\)-äquivalentes Messmodell
Paralleles Messmodell
Censor_L_parallel <- "
Censor =~ l * Censor_1 + l * Censor_2 + l * Censor_3 + l * Censor_4 # gleiche Ladung l für alle Items
Censor_1 ~~ v * Censor_1 # gleiche Varianz v für alle Items
Censor_2 ~~ v * Censor_2 # gleiche Varianz v für alle Items
Censor_3 ~~ v * Censor_3 # gleiche Varianz v für alle Items
Censor_4 ~~ v * Censor_4 # gleiche Varianz v für alle Items
"| lhs | op | rhs | est | se | z | pvalue | ci.lower | ci.upper | std.all |
|---|---|---|---|---|---|---|---|---|---|
| Censor | =~ | Censor_1 | 1.53 | 0.06 | 25.14 | 0.00 | 1.41 | 1.65 | 0.92 |
| Censor | =~ | Censor_2 | 1.55 | 0.06 | 24.74 | 0.00 | 1.43 | 1.67 | 0.97 |
| Censor | =~ | Censor_3 | 1.58 | 0.06 | 25.01 | 0.00 | 1.45 | 1.70 | 0.96 |
| Censor | =~ | Censor_4 | 1.62 | 0.06 | 27.10 | 0.00 | 1.50 | 1.73 | 0.95 |
| Censor_1 | Censor_1 | 0.41 | 0.09 | 4.72 | 0.00 | 0.24 | 0.57 | 0.15 | |
| Censor_2 | Censor_2 | 0.17 | 0.04 | 4.02 | 0.00 | 0.09 | 0.25 | 0.07 | |
| Censor_3 | Censor_3 | 0.19 | 0.04 | 4.70 | 0.00 | 0.11 | 0.27 | 0.07 | |
| Censor_4 | Censor_4 | 0.26 | 0.06 | 4.39 | 0.00 | 0.14 | 0.37 | 0.09 | |
| Censor | Censor | 1.00 | 0.00 | 1.00 | 1.00 | 1.00 |
| lhs | op | rhs | label | est | se | z | pvalue | ci.lower | ci.upper | std.all |
|---|---|---|---|---|---|---|---|---|---|---|
| Censor | =~ | Censor_1 | l | 1.57 | 0.06 | 27.01 | 0.00 | 1.46 | 1.68 | 0.93 |
| Censor | =~ | Censor_2 | l | 1.57 | 0.06 | 27.01 | 0.00 | 1.46 | 1.68 | 0.97 |
| Censor | =~ | Censor_3 | l | 1.57 | 0.06 | 27.01 | 0.00 | 1.46 | 1.68 | 0.96 |
| Censor | =~ | Censor_4 | l | 1.57 | 0.06 | 27.01 | 0.00 | 1.46 | 1.68 | 0.95 |
| Censor_1 | Censor_1 | 0.40 | 0.09 | 4.70 | 0.00 | 0.23 | 0.57 | 0.14 | ||
| Censor_2 | Censor_2 | 0.16 | 0.04 | 4.16 | 0.00 | 0.09 | 0.24 | 0.06 | ||
| Censor_3 | Censor_3 | 0.20 | 0.04 | 4.88 | 0.00 | 0.12 | 0.28 | 0.07 | ||
| Censor_4 | Censor_4 | 0.27 | 0.06 | 4.49 | 0.00 | 0.15 | 0.38 | 0.10 | ||
| Censor | Censor | 1.00 | 0.00 | 1.00 | 1.00 | 1.00 |
| lhs | op | rhs | label | est | se | z | pvalue | ci.lower | ci.upper | std.all |
|---|---|---|---|---|---|---|---|---|---|---|
| Censor | =~ | Censor_1 | l | 1.57 | 0.06 | 27.45 | 0.00 | 1.46 | 1.68 | 0.95 |
| Censor | =~ | Censor_2 | l | 1.57 | 0.06 | 27.45 | 0.00 | 1.46 | 1.68 | 0.95 |
| Censor | =~ | Censor_3 | l | 1.57 | 0.06 | 27.45 | 0.00 | 1.46 | 1.68 | 0.95 |
| Censor | =~ | Censor_4 | l | 1.57 | 0.06 | 27.45 | 0.00 | 1.46 | 1.68 | 0.95 |
| Censor_1 | Censor_1 | v | 0.26 | 0.03 | 7.93 | 0.00 | 0.19 | 0.32 | 0.10 | |
| Censor_2 | Censor_2 | v | 0.26 | 0.03 | 7.93 | 0.00 | 0.19 | 0.32 | 0.10 | |
| Censor_3 | Censor_3 | v | 0.26 | 0.03 | 7.93 | 0.00 | 0.19 | 0.32 | 0.10 | |
| Censor_4 | Censor_4 | v | 0.26 | 0.03 | 7.93 | 0.00 | 0.19 | 0.32 | 0.10 | |
| Censor | Censor | 1.00 | 0.00 | 1.00 | 1.00 | 1.00 |
Kongenerisch

\(\tau\)-äquivalent

Parallel

################### Nested Model Comparison #########################
Scaled Chi-Squared Difference Test (method = "satorra.bentler.2001")
lavaan->unknown():
lavaan NOTE: The "Chisq" column contains standard test statistics, not the robust test that should be reported per model. A robust difference test is a
function of two standard (not robust) statistics.
Df AIC BIC Chisq Chisq diff RMSEA Df diff Pr(>Chisq)
Censor_L_fit 2 3769.2 3801.1 15.756
Censor_L_tau_fit 5 3768.9 3788.8 21.432 5.5385 0.046737 3 0.136354
Censor_L_parallel_fit 8 3817.7 3825.6 76.219 12.8183 0.187709 3 0.005046 **
---
Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
####################### Model Fit Indices ###########################
rmsea.robust srmr cfi.robust
Censor_L_fit .102 .006† 0.996†
Censor_L_tau_fit .074† .028 0.995
Censor_L_parallel_fit .129 .031 .977
################## Differences in Fit Indices #######################
rmsea.robust srmr cfi.robust
Censor_L_tau_fit - Censor_L_fit -0.028 0.022 -0.001
Censor_L_parallel_fit - Censor_L_tau_fit 0.055 0.003 -0.018
Wichtig unter anderem bei ländervergleichenden (Messen meine Instrumente in allen Sprachen/Kulturen dasselbe?), altersvergleichenden (Verstehen alle Altersgruppe die Items gleich?) oder experimentellen (Verändert das Treatment die Messung des Outcome?) Studien.
Obwohl alle verstehen, warum das wichtig wäre, wird es viel zu selten geprüft.
Welche Formen der Invarianz wir annehmen dürfen, wird auf Basis eines Modellvergleichs bestimmt.
Beispiel: Messinvarianz nach Treatment (experience, facts)
Facts
| lhs | op | rhs | group | est | ci.lower | ci.upper | std.all |
|---|---|---|---|---|---|---|---|
| Censor | =~ | Censor_1 | 1 | 1.63 | 1.47 | 1.79 | 0.92 |
| Censor | =~ | Censor_2 | 1 | 1.68 | 1.51 | 1.84 | 0.97 |
| Censor | =~ | Censor_3 | 1 | 1.73 | 1.57 | 1.90 | 0.95 |
| Censor | =~ | Censor_4 | 1 | 1.73 | 1.58 | 1.89 | 0.94 |
| Censor_1 | ~1 | 1 | 3.07 | 2.82 | 3.32 | 1.73 | |
| Censor_2 | ~1 | 1 | 2.86 | 2.62 | 3.11 | 1.65 | |
| Censor_3 | ~1 | 1 | 2.93 | 2.68 | 3.19 | 1.61 | |
| Censor_4 | ~1 | 1 | 3.04 | 2.78 | 3.30 | 1.65 | |
| Censor | ~1 | 1 | 0.00 | 0.00 | 0.00 | 0.00 |
Experience
| lhs | op | rhs | group | est | ci.lower | ci.upper | std.all |
|---|---|---|---|---|---|---|---|
| Censor | =~ | Censor_1 | 2 | 1.33 | 1.16 | 1.50 | 0.92 |
| Censor | =~ | Censor_2 | 2 | 1.35 | 1.18 | 1.52 | 0.97 |
| Censor | =~ | Censor_3 | 2 | 1.31 | 1.14 | 1.48 | 0.97 |
| Censor | =~ | Censor_4 | 2 | 1.38 | 1.21 | 1.56 | 0.97 |
| Censor_1 | ~1 | 2 | 2.34 | 2.14 | 2.54 | 1.62 | |
| Censor_2 | ~1 | 2 | 2.24 | 2.04 | 2.43 | 1.61 | |
| Censor_3 | ~1 | 2 | 2.24 | 2.05 | 2.42 | 1.66 | |
| Censor_4 | ~1 | 2 | 2.26 | 2.06 | 2.46 | 1.58 | |
| Censor | ~1 | 2 | 0.00 | 0.00 | 0.00 | 0.00 |
Facts
| lhs | op | rhs | group | est | ci.lower | ci.upper | std.all |
|---|---|---|---|---|---|---|---|
| Censor | =~ | Censor_1 | 1 | 1.66 | 1.51 | 1.80 | 0.92 |
| Censor | =~ | Censor_2 | 1 | 1.69 | 1.54 | 1.84 | 0.97 |
| Censor | =~ | Censor_3 | 1 | 1.68 | 1.53 | 1.84 | 0.95 |
| Censor | =~ | Censor_4 | 1 | 1.74 | 1.59 | 1.90 | 0.94 |
| Censor_1 | ~1 | 1 | 3.07 | 2.82 | 3.32 | 1.71 | |
| Censor_2 | ~1 | 1 | 2.86 | 2.62 | 3.11 | 1.64 | |
| Censor_3 | ~1 | 1 | 2.93 | 2.68 | 3.19 | 1.65 | |
| Censor_4 | ~1 | 1 | 3.04 | 2.78 | 3.30 | 1.64 | |
| Censor | ~1 | 1 | 0.00 | 0.00 | 0.00 | 0.00 |
Experience
| lhs | op | rhs | group | est | ci.lower | ci.upper | std.all |
|---|---|---|---|---|---|---|---|
| Censor | =~ | Censor_1 | 2 | 1.66 | 1.51 | 1.80 | 0.92 |
| Censor | =~ | Censor_2 | 2 | 1.69 | 1.54 | 1.84 | 0.97 |
| Censor | =~ | Censor_3 | 2 | 1.68 | 1.53 | 1.84 | 0.97 |
| Censor | =~ | Censor_4 | 2 | 1.74 | 1.59 | 1.90 | 0.97 |
| Censor_1 | ~1 | 2 | 2.34 | 2.14 | 2.54 | 1.64 | |
| Censor_2 | ~1 | 2 | 2.24 | 2.04 | 2.43 | 1.62 | |
| Censor_3 | ~1 | 2 | 2.24 | 2.05 | 2.42 | 1.64 | |
| Censor_4 | ~1 | 2 | 2.26 | 2.06 | 2.46 | 1.58 | |
| Censor | ~1 | 2 | 0.00 | 0.00 | 0.00 | 0.00 |
Facts
| lhs | op | rhs | group | est | ci.lower | ci.upper | std.all |
|---|---|---|---|---|---|---|---|
| Censor | =~ | Censor_1 | 1 | 1.67 | 1.52 | 1.81 | 0.92 |
| Censor | =~ | Censor_2 | 1 | 1.68 | 1.53 | 1.83 | 0.97 |
| Censor | =~ | Censor_3 | 1 | 1.69 | 1.53 | 1.84 | 0.95 |
| Censor | =~ | Censor_4 | 1 | 1.75 | 1.60 | 1.91 | 0.94 |
| Censor_1 | ~1 | 1 | 3.04 | 2.80 | 3.28 | 1.69 | |
| Censor_2 | ~1 | 1 | 2.90 | 2.66 | 3.15 | 1.67 | |
| Censor_3 | ~1 | 1 | 2.93 | 2.68 | 3.18 | 1.65 | |
| Censor_4 | ~1 | 1 | 2.99 | 2.74 | 3.25 | 1.61 | |
| Censor | ~1 | 1 | 0.00 | 0.00 | 0.00 | 0.00 |
Experience
| lhs | op | rhs | group | est | ci.lower | ci.upper | std.all |
|---|---|---|---|---|---|---|---|
| Censor | =~ | Censor_1 | 2 | 1.67 | 1.52 | 1.81 | 0.92 |
| Censor | =~ | Censor_2 | 2 | 1.68 | 1.53 | 1.83 | 0.97 |
| Censor | =~ | Censor_3 | 2 | 1.69 | 1.53 | 1.84 | 0.97 |
| Censor | =~ | Censor_4 | 2 | 1.75 | 1.60 | 1.91 | 0.97 |
| Censor_1 | ~1 | 2 | 3.04 | 2.80 | 3.28 | 2.13 | |
| Censor_2 | ~1 | 2 | 2.90 | 2.66 | 3.15 | 2.11 | |
| Censor_3 | ~1 | 2 | 2.93 | 2.68 | 3.18 | 2.14 | |
| Censor_4 | ~1 | 2 | 2.99 | 2.74 | 3.25 | 2.09 | |
| Censor | ~1 | 2 | -0.41 | -0.58 | -0.24 | -0.52 |
################### Nested Model Comparison #########################
Scaled Chi-Squared Difference Test (method = "satorra.bentler.2001")
lavaan->unknown():
lavaan NOTE: The "Chisq" column contains standard test statistics, not the robust test that should be reported per model. A robust difference test is a
function of two standard (not robust) statistics.
Df AIC BIC Chisq Chisq diff RMSEA Df diff Pr(>Chisq)
Censor_L_config_fit 4 3674.5 3770.2 16.368
Censor_L_metric_fit 7 3671.9 3755.5 19.689 3.2173 0.019407 3 0.3593
Censor_L_scalar_fit 10 3670.8 3742.6 24.668 4.7468 0.055470 3 0.1913
####################### Model Fit Indices ###########################
rmsea.robust srmr cfi.robust
Censor_L_config_fit .043 .005† 0.999†
Censor_L_metric_fit .035† .018 0.999
Censor_L_scalar_fit .042 .021 0.998
################## Differences in Fit Indices #######################
rmsea.robust srmr cfi.robust
Censor_L_metric_fit - Censor_L_config_fit -0.008 0.013 0.000
Censor_L_scalar_fit - Censor_L_metric_fit 0.007 0.003 -0.001
Harm: What Morgan tells others on Twitter …
Harm_1)Harm_2)Harm_3)Lie: What Morgan tells others on Twitter …
Lie_1)Lie_2)Lie_3)1–7, strongly disagree to strongly agree.

| lhs | op | rhs | est | se | z | pvalue | ci.lower | ci.upper | std.all |
|---|---|---|---|---|---|---|---|---|---|
| Harm | =~ | Harm_1 | 1.68 | 0.05 | 33.64 | 0 | 1.59 | 1.78 | 0.96 |
| Harm | =~ | Harm_2 | 1.72 | 0.05 | 35.28 | 0 | 1.62 | 1.81 | 0.96 |
| Harm | =~ | Harm_3 | 1.62 | 0.06 | 27.85 | 0 | 1.51 | 1.73 | 0.91 |
| Lie | =~ | Lie_1 | 1.62 | 0.05 | 31.67 | 0 | 1.52 | 1.72 | 0.96 |
| Lie | =~ | Lie_2 | 1.59 | 0.06 | 28.96 | 0 | 1.49 | 1.70 | 0.93 |
| Lie | =~ | Lie_3 | 1.57 | 0.06 | 28.60 | 0 | 1.47 | 1.68 | 0.93 |
| Harm | Lie | 0.59 | 0.04 | 13.57 | 0 | 0.51 | 0.68 | 0.59 |
$Harm
Composite `Harm` is composed of observed variables:
Harm_1, Harm_2, Harm_3
True-score variance is represented by common factor(s):
Harm
Total variance of composite `Harm` determined from the unrestricted model.
The proportion attributable to "true" scores is its model-based estimate of reliability ("omega"):
[1] 0.958
$Lie
Composite `Lie` is composed of observed variables:
Lie_1, Lie_2, Lie_3
True-score variance is represented by common factor(s):
Lie
Total variance of composite `Lie` determined from the unrestricted model.
The proportion attributable to "true" scores is its model-based estimate of reliability ("omega"):
[1] 0.957
chisq.scaled 14.072
df.scaled 8.000
pvalue.scaled 0.080
rmsea.robust 0.048
rmsea.ci.lower.robust 0.000
rmsea.ci.upper.robust 0.089
srmr 0.014
cfi.robust 0.997
Your DFI cutoffs:
SRMR RMSEA CFI Magnitude
Level-0 0.018 0.049 0.997 NONE
Specificity 95% 95% 95%
Level-1 0.02 0.07 0.995 0.138
Sensitivity 95% 95% 95%
Empirical fit indices:
Chi-Square df p-value SRMR RMSEA CFI
14.072 8 0.08 0.014 0.048 0.997
Notes:
-'Sensitivity' is % of hypothetically misspecified models correctly identified by cutoff in DFI simulation
-Cutoffs with 95% sensitivity are reported when possible
-If sensitivity is <50%, cutoffs will be supressed
chisq.scaled 0
df.scaled 0
pvalue.scaled NA
rmsea.robust 0
rmsea.ci.lower.robust 0
rmsea.ci.upper.robust 0
srmr 0
cfi.robust NA
| lhs | op | rhs | mi | epc | sepc.lv | sepc.all | sepc.nox |
|---|---|---|---|---|---|---|---|
| Lie_1 | Lie_3 | 6.18 | 0.21 | 0.21 | 0.68 | 0.68 | |
| Harm | =~ | Lie_2 | 6.18 | 0.12 | 0.12 | 0.07 | 0.07 |
| Harm_2 | Lie_3 | 5.74 | 0.06 | 0.06 | 0.18 | 0.18 | |
| Harm_2 | Lie_1 | 5.67 | -0.05 | -0.05 | -0.20 | -0.20 | |
| Harm_1 | Lie_3 | 4.26 | -0.05 | -0.05 | -0.15 | -0.15 | |
| Harm_3 | Lie_1 | 3.21 | 0.05 | 0.05 | 0.13 | 0.13 |
Lie_2 (“is wrong”) würde gerne auch etwas zur latenten Variable Harm beitragen.mediation_I <- "
# Modelle
# Mediator 1: Harm
M_Harm ~ a1 * Condition
# Mediator 2: False
M_Lie ~ a2 * Condition
# aV: Endorsement of Censorship
M_Censor ~ c_ * Condition + b1 * M_Harm + b2 * M_Lie
# Kovarianz zwischen Mediatoren
M_Harm ~~ M_Lie
# Effekte
# Indirekte Effekte
indirect_harm := a1 * b1
indirect_lie := a2 * b2
# Direkter Effekt
direct := c_
# Totaler Effekt
total := direct + indirect_harm + indirect_lie
"| lhs | op | rhs | label | est | se | z | pvalue | ci.lower | ci.upper | std.all |
|---|---|---|---|---|---|---|---|---|---|---|
| M_Harm | ~ | Condition | a1 | -0.81 | 0.17 | -4.89 | 0.00 | -1.14 | -0.49 | -0.24 |
| M_Lie | ~ | Condition | a2 | -1.55 | 0.14 | -10.77 | 0.00 | -1.83 | -1.27 | -0.48 |
| M_Censor | ~ | Condition | c_ | -0.20 | 0.17 | -1.23 | 0.22 | -0.53 | 0.12 | -0.06 |
| M_Censor | ~ | M_Harm | b1 | 0.21 | 0.05 | 4.09 | 0.00 | 0.11 | 0.31 | 0.23 |
| M_Censor | ~ | M_Lie | b2 | 0.22 | 0.06 | 3.45 | 0.00 | 0.09 | 0.34 | 0.22 |
| indirect_harm | := | a1*b1 | indirect_harm | -0.17 | 0.05 | -3.16 | 0.00 | -0.28 | -0.07 | -0.05 |
| indirect_lie | := | a2*b2 | indirect_lie | -0.33 | 0.10 | -3.27 | 0.00 | -0.53 | -0.13 | -0.11 |
| direct | := | c_ | direct | -0.20 | 0.17 | -1.23 | 0.22 | -0.53 | 0.12 | -0.06 |
| total | := | direct+indirect_harm+indirect_lie | total | -0.71 | 0.16 | -4.57 | 0.00 | -1.02 | -0.41 | -0.22 |
Einfaches Modell ohne Restriktionen; ginge besser, siehe oben.
mediation_L <- "
# Konstrukte
Harm =~ Harm_1 + Harm_2 + Harm_3
Lie =~ Lie_1 + Lie_2 + Lie_3
Censor =~ Censor_1 + Censor_2 + Censor_3 + Censor_4
# Korrelation zwischen Konstrukten
Harm ~~ Lie
# Modelle
# Mediator 1: Harm
Harm ~ a1 * Condition
# Mediator 2: False
Lie ~ a2 * Condition
# aV: Endorsement of Censorship
Censor ~ c_ * Condition + b1 * Harm + b2 * Lie
# Effekte
# Indirekte Effekte
indirect_harm := a1 * b1
indirect_lie := a2 * b2
# Direkter Effekt
direct := c_
# Totaler Effekt
total := direct + indirect_harm + indirect_lie
"| lhs | op | rhs | label | est | se | z | pvalue | ci.lower | ci.upper | std.all |
|---|---|---|---|---|---|---|---|---|---|---|
| Harm | =~ | Harm_1 | 1.00 | 0.00 | 1.00 | 1.00 | 0.96 | |||
| Harm | =~ | Harm_2 | 1.02 | 0.02 | 49.05 | 0.00 | 0.98 | 1.06 | 0.96 | |
| Harm | =~ | Harm_3 | 0.96 | 0.03 | 37.11 | 0.00 | 0.91 | 1.01 | 0.91 | |
| Lie | =~ | Lie_1 | 1.00 | 0.00 | 1.00 | 1.00 | 0.96 | |||
| Lie | =~ | Lie_2 | 0.98 | 0.02 | 44.56 | 0.00 | 0.94 | 1.03 | 0.93 | |
| Lie | =~ | Lie_3 | 0.97 | 0.02 | 47.58 | 0.00 | 0.93 | 1.01 | 0.93 | |
| Censor | =~ | Censor_1 | 1.00 | 0.00 | 1.00 | 1.00 | 0.92 | |||
| Censor | =~ | Censor_2 | 1.01 | 0.02 | 40.72 | 0.00 | 0.96 | 1.06 | 0.97 | |
| Censor | =~ | Censor_3 | 1.03 | 0.03 | 34.33 | 0.00 | 0.97 | 1.09 | 0.96 | |
| Censor | =~ | Censor_4 | 1.06 | 0.03 | 38.01 | 0.00 | 1.00 | 1.11 | 0.95 | |
| Harm | ~ | Condition | a1 | -0.83 | 0.17 | -4.95 | 0.00 | -1.16 | -0.50 | -0.25 |
| Lie | ~ | Condition | a2 | -1.57 | 0.15 | -10.72 | 0.00 | -1.86 | -1.29 | -0.49 |
| Censor | ~ | Condition | c_ | -0.18 | 0.16 | -1.07 | 0.28 | -0.50 | 0.15 | -0.06 |
| Censor | ~ | Harm | b1 | 0.23 | 0.06 | 4.04 | 0.00 | 0.12 | 0.34 | 0.25 |
| Censor | ~ | Lie | b2 | 0.20 | 0.07 | 3.03 | 0.00 | 0.07 | 0.33 | 0.21 |
| indirect_harm | := | a1*b1 | indirect_harm | -0.19 | 0.06 | -3.16 | 0.00 | -0.31 | -0.07 | -0.06 |
| indirect_lie | := | a2*b2 | indirect_lie | -0.32 | 0.11 | -2.91 | 0.00 | -0.53 | -0.10 | -0.10 |
| direct | := | c_ | direct | -0.18 | 0.16 | -1.07 | 0.28 | -0.50 | 0.15 | -0.06 |
| total | := | direct+indirect_harm+indirect_lie | total | -0.68 | 0.15 | -4.49 | 0.00 | -0.98 | -0.38 | -0.22 |
Mittelwertindices
| label | est | ci.lower | ci.upper | std.all |
|---|---|---|---|---|
| a1 | -0.81 | -1.14 | -0.49 | -0.24 |
| a2 | -1.55 | -1.83 | -1.27 | -0.48 |
| c_ | -0.20 | -0.53 | 0.12 | -0.06 |
| b1 | 0.21 | 0.11 | 0.31 | 0.23 |
| b2 | 0.22 | 0.09 | 0.34 | 0.22 |
| indirect_harm | -0.17 | -0.28 | -0.07 | -0.05 |
| indirect_lie | -0.33 | -0.53 | -0.13 | -0.11 |
| direct | -0.20 | -0.53 | 0.12 | -0.06 |
| total | -0.71 | -1.02 | -0.41 | -0.22 |
Latente Variablen
| label | est | ci.lower | ci.upper | std.all |
|---|---|---|---|---|
| a1 | -0.83 | -1.16 | -0.50 | -0.25 |
| a2 | -1.57 | -1.86 | -1.29 | -0.49 |
| c_ | -0.18 | -0.50 | 0.15 | -0.06 |
| b1 | 0.23 | 0.12 | 0.34 | 0.25 |
| b2 | 0.20 | 0.07 | 0.33 | 0.21 |
| indirect_harm | -0.19 | -0.31 | -0.07 | -0.06 |
| indirect_lie | -0.32 | -0.53 | -0.10 | -0.10 |
| direct | -0.18 | -0.50 | 0.15 | -0.06 |
| total | -0.68 | -0.98 | -0.38 | -0.22 |
Noch nicht entschieden
Marko Bachl