ellenbogen within-subject
2023-01-20
Description
- goal was to pilot a within-subject version of the TI task to check time-dependent consolidation effect on inference performance
- 73 subs collected via Pavlovia
- Setup: learning/feedback terminated 66% in all 3 versions of the
pilot, what did vary was a) stimulus set b) number of blocks of trials
shown to participant without feedback
- version 1: facescene - counterbalanced stimuli of faces or scenes, learning regime: learning terminates upon reaching criteria (66%) >> unbalanced learning trials between conditions and unbalanced presentation. Showed main-effect of time for Inference performance
- version 2 (idea: try to use 1 stimuli set for better control): objects - counterbalanced stimuli of objects, learning regime: learning terminates upon reaching criteria (66%), but continues to present stimuli without feedback until 9th block >> unbalanced learning trials between H1 and H2, but balanced presentation between H1 and H2. Didn’t show main-effect of time for Inference performance
- version 3: scenes - counterbalanced stimuli of scenes, learning regime: learning terminates upon reaching criteria (66%) for H1, H2 learning fixed in accordance to H1 blockcount, >> balanced learning trials H1 and H2 and balanced presentation between H1 and H2, but unbalanced performance
library(afex) ## to run anova## Loading required package: lme4## Loading required package: Matrix##
## Attaching package: 'Matrix'## The following objects are masked from 'package:tidyr':
##
## expand, pack, unpack## ************
## Welcome to afex. For support visit: http://afex.singmann.science/## - Functions for ANOVAs: aov_car(), aov_ez(), and aov_4()
## - Methods for calculating p-values with mixed(): 'S', 'KR', 'LRT', and 'PB'
## - 'afex_aov' and 'mixed' objects can be passed to emmeans() for follow-up tests
## - NEWS: emmeans() for ANOVA models now uses model = 'multivariate' as default.
## - Get and set global package options with: afex_options()
## - Set orthogonal sum-to-zero contrasts globally: set_sum_contrasts()
## - For example analyses see: browseVignettes("afex")
## ************##
## Attaching package: 'afex'## The following object is masked from 'package:lme4':
##
## lmerlibrary(ggstatsplot)
#73 subs collected thru pavlovia
summary.all %>% ungroup() %>% summarise(n_distinct(participant))## # A tibble: 1 × 1
## `n_distinct(participant)`
## <int>
## 1 73grouped_ggwithinstats(
data = session3.pretty.filtered.facescene %>%
mutate(across(pairType, recode, '24'= "inf", '25'= "inf", '35'="inf",'16'= "anchor")) %>% group_by(participant,pairType, awareness) %>% summarise(H1 = mean(H1), H2 = mean(H2)) %>% gather(hierarchy, meanPerf, c("H1","H2")),
subject.id = participant,
x = hierarchy,
y = meanPerf,
type = "parametric",
xlab = "Condition",
grouping.var = pairType,
outlier.label = NULL,
bf.message = FALSE,
annotation.args = list(
title = "Version 1"))## `summarise()` has grouped output by 'participant', 'pairType'. You can override
## using the `.groups` argument.
grouped_ggwithinstats(
data = rbind(session3.pretty.filtered.shape,session3.pretty.filtered.noshape) %>%
mutate(across(pairType, recode, '24'= "inf", '25'= "inf", '35'="inf",'16'= "anchor")) %>% group_by(participant,pairType, awareness) %>% summarise(H1 = mean(H1), H2 = mean(H2)) %>% gather(hierarchy, meanPerf, c("H1","H2")),
subject.id = participant,
x = hierarchy,
y = meanPerf,
type = "parametric",
xlab = "Condition",
grouping.var = pairType,
outlier.label = NULL,
bf.message = FALSE,annotation.args = list(
title = "Version 3"))## `summarise()` has grouped output by 'participant', 'pairType'. You can override
## using the `.groups` argument.
grouped_ggwithinstats(
data = session3.pretty.filtered.scenes %>%
mutate(across(pairType, recode, '24'= "inf", '25'= "inf", '35'="inf",'16'= "anchor")) %>% group_by(participant,pairType, awareness) %>% summarise(H1 = mean(H1), H2 = mean(H2)) %>% gather(hierarchy, meanPerf, c("H1","H2")),
subject.id = participant,
x = hierarchy,
y = meanPerf,
type = "parametric",
xlab = "Condition",
grouping.var = pairType,
outlier.label = NULL,
bf.message = FALSE,
annotation.args = list(
title = "Version 2"))## `summarise()` has grouped output by 'participant', 'pairType'. You can override
## using the `.groups` argument.
LMM
- meanPerf.y = Immediate (session-specific) Premise pair performance
- sig main effect of Hierarchy (H1: Day 1 session, H2: Day 2 session) and Immediate Premise pair performance on Inference performance (at session 3).
load("awareness_score_scenes.Rdata")
load("awareness_score_facescene.Rdata")
load("awareness_score_shape.Rdata")
load("awareness_score_obj.Rdata")
load("session3_pretty_filtered_obj.Rdata")
lmm.scenes <- bind_rows(session3.pretty.filtered.scenes %>%
mutate(across(pairType, dplyr::recode, '24'= "inf", '25'= "inf", '35'="inf",'16'= "anchor")) %>% group_by(participant,pairType, awareness) %>% summarise(H1 = mean(H1), H2 = mean(H2)) %>% gather(hierarchy, meanPerf, c("H1","H2")) %>% mutate(session=3) ,rbind(session1.avgs.scenes, session2.avgs.scenes) %>% filter(participant %in% session3.pretty.filtered.scenes$participant) %>% group_by(participant,session,blockcount) %>% summarise(premise=mean(key_resp_im.corr)) %>% gather(pairType, meanPerf, premise) %>% mutate(hierarchy=paste0("H",session))) %>% mutate(stimuli="scene") %>% select(-awareness) %>% left_join(awareness_score.scenes, c("participant"="code")) %>% mutate(participant=paste0("scn",participant), timeelapsed = timeelapsed*60)## `summarise()` has grouped output by 'participant', 'pairType'. You can override
## using the `.groups` argument.
## `summarise()` has grouped output by 'participant', 'session'. You can override
## using the `.groups` argument.lmm.facescene <- bind_rows(session3.pretty.filtered.facescene %>%
mutate(across(pairType, dplyr::recode, '24'= "inf", '25'= "inf", '35'="inf",'16'= "anchor")) %>% group_by(participant,pairType, awareness) %>% summarise(H1 = mean(H1), H2 = mean(H2)) %>% gather(hierarchy, meanPerf, c("H1","H2")) %>% mutate(session=3), rbind(session1.avgs.facescene, session2.avgs.facescene) %>% filter(participant %in% session3.pretty.filtered.facescene$participant) %>% group_by(participant,session,blockcount) %>% summarise(premise=mean(key_resp_im.corr)) %>% gather(pairType, meanPerf, premise) %>% mutate(hierarchy=paste0("H",session))) %>% mutate(stimuli="facescene") %>% select(-awareness) %>% left_join(awareness_score.facescene, c("participant"="code")) %>% mutate(participant=paste0("fsc",participant))## `summarise()` has grouped output by 'participant', 'pairType'. You can override
## using the `.groups` argument.
## `summarise()` has grouped output by 'participant', 'session'. You can override
## using the `.groups` argument.lmm.shape <- bind_rows(session3.pretty.filtered.shape %>%
mutate(across(pairType, dplyr::recode, '24'= "inf", '25'= "inf", '35'="inf",'16'= "anchor")) %>% group_by(participant,pairType, awareness) %>% summarise(H1 = mean(H1), H2 = mean(H2)) %>% gather(hierarchy, meanPerf, c("H1","H2")) %>% mutate(session=3), rbind(session1.avgs.shape, session2.avgs.shape) %>% filter(participant %in% session3.pretty.filtered.shape$participant) %>% group_by(participant,session,blockcount) %>% summarise(premise=mean(key_resp_im.corr)) %>% gather(pairType, meanPerf, premise) %>% mutate(hierarchy=paste0("H",session))) %>% mutate(stimuli="obj") %>% select(-awareness) %>% left_join(awareness_score.shape, c("participant"="code")) %>% mutate(participant=paste0("o_s",participant))## `summarise()` has grouped output by 'participant', 'pairType'. You can override
## using the `.groups` argument.
## `summarise()` has grouped output by 'participant', 'session'. You can override
## using the `.groups` argument.# save(session3.pretty.filtered.obj, file = "session3_pretty_filtered_obj.Rdata")
#load("session3_pretty_filtered_obj.Rdata")
lmm.noshape <- bind_rows(session3.pretty.filtered.obj %>%
mutate(across(pairType, dplyr::recode, '24'= "inf", '25'= "inf", '35'="inf",'16'= "anchor")) %>% group_by(participant,pairType, awareness) %>% summarise(H1 = mean(H1), H2 = mean(H2)) %>% gather(hierarchy, meanPerf, c("H1","H2")) %>% mutate(session=3), rbind(session1.avgs.obj, session2.avgs.obj) %>% filter(participant %in% session3.pretty.filtered.obj$participant) %>% group_by(participant,session,blockcount) %>% summarise(premise=mean(key_resp_im.corr)) %>% gather(pairType, meanPerf, premise) %>% mutate(hierarchy=paste0("H",session))) %>% mutate(stimuli="obj") %>% select(-awareness) %>% left_join(awareness_score.obj, c("participant"="code")) %>% mutate(participant=paste0("o_ns",participant))## `summarise()` has grouped output by 'participant', 'pairType'. You can override
## using the `.groups` argument.
## `summarise()` has grouped output by 'participant', 'session'. You can override
## using the `.groups` argument.lmm.combined <- bind_rows(lmm.scenes,lmm.facescene,lmm.shape,lmm.noshape)
s1s2.im <- lmm.combined %>% filter(session %in% c(1,2))
lmm.combined.wide <- lmm.combined %>% filter(session == 3) %>% left_join(.,s1s2.im,by=c("participant"="participant","hierarchy"="hierarchy"))
glimpse(lmm.combined)## Rows: 584
## Columns: 9
## Groups: participant, pairType [219]
## $ participant <chr> "scnBIHKJQY", "scnBIHKJQY", "scnBIHKJQY", "scnGYEDWMN", "s…
## $ pairType <chr> "anchor", "inf", "premise", "anchor", "inf", "premise", "a…
## $ hierarchy <chr> "H1", "H1", "H1", "H1", "H1", "H1", "H1", "H1", "H1", "H1"…
## $ meanPerf <dbl> 0.8750000, 0.1250000, 0.5750000, 0.6250000, 0.3333333, 0.4…
## $ session <dbl> 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3…
## $ blockcount <dbl> NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA…
## $ stimuli <chr> "scene", "scene", "scene", "scene", "scene", "scene", "sce…
## $ awareness <dbl> 8, 8, 8, 8, 8, 8, 3, 3, 3, 5, 5, 5, 11, 11, 11, 7, 7, 7, 8…
## $ timeelapsed <dbl> 186.9333, 186.9333, 186.9333, 204.5167, 204.5167, 204.5167…library(lme4)
library(lmerTest)##
## Attaching package: 'lmerTest'
##
## The following object is masked from 'package:lme4':
##
## lmer
##
## The following object is masked from 'package:stats':
##
## step# new attempt : with wide
m1 <- lmer(meanPerf.x ~ 1 + (1 | participant), data = lmm.combined.wide %>% filter(pairType.x == "inf"), REML=FALSE)
m2 <- lmer(meanPerf.x ~ hierarchy + meanPerf.y + (1 | participant), data = lmm.combined.wide %>% filter(pairType.x == "inf"))
m3 <- lmer(meanPerf.x ~ hierarchy * meanPerf.y + (1 | participant) , data = lmm.combined.wide %>% filter(pairType.x == "inf"))
tab_model(m1,m2,m3)| meanPerf.x | meanPerf.x | meanPerf.x | |||||||
|---|---|---|---|---|---|---|---|---|---|
| Predictors | Estimates | CI | p | Estimates | CI | p | Estimates | CI | p |
| (Intercept) | 0.54 | 0.49 – 0.59 | <0.001 | 0.20 | -0.02 – 0.43 | 0.072 | 0.05 | -0.22 – 0.33 | 0.696 |
| hierarchy [H2] | -0.08 | -0.16 – 0.00 | 0.051 | 0.28 | -0.13 – 0.68 | 0.179 | |||
| meanPerf y | 0.50 | 0.20 – 0.79 | 0.001 | 0.71 | 0.34 – 1.08 | <0.001 | |||
|
hierarchy [H2] × meanPerf y |
-0.46 | -0.98 – 0.05 | 0.078 | ||||||
| Random Effects | |||||||||
| σ2 | 0.06 | 0.05 | 0.05 | ||||||
| τ00 | 0.01 participant | 0.02 participant | 0.02 participant | ||||||
| ICC | 0.18 | 0.29 | 0.32 | ||||||
| N | 73 participant | 73 participant | 73 participant | ||||||
| Observations | 146 | 146 | 146 | ||||||
| Marginal R2 / Conditional R2 | 0.000 / 0.180 | 0.077 / 0.342 | 0.094 / 0.380 | ||||||
plot_model(m2, type="pred", terms = c("meanPerf.y","hierarchy"))
# new attempt : with wide; without below chance immediate premise performers // affects two participants
m1 <- lmer(meanPerf.x ~ 1 + (1 | participant), data = lmm.combined.wide %>% filter(pairType.x == "inf", meanPerf.y>0.5), REML=FALSE)
m2 <- lmer(meanPerf.x ~ hierarchy + meanPerf.y + (1 | participant), data = lmm.combined.wide %>% filter(pairType.x == "inf",meanPerf.y>0.5))
m3 <- lmer(meanPerf.x ~ hierarchy * meanPerf.y + (1 | participant) , data = lmm.combined.wide %>% filter(pairType.x == "inf",meanPerf.y>0.5))
anova(m1,m2,m3)## refitting model(s) with ML (instead of REML)## Data: lmm.combined.wide %>% filter(pairType.x == "inf", meanPerf.y > ...
## Models:
## m1: meanPerf.x ~ 1 + (1 | participant)
## m2: meanPerf.x ~ hierarchy + meanPerf.y + (1 | participant)
## m3: meanPerf.x ~ hierarchy * meanPerf.y + (1 | participant)
## npar AIC BIC logLik deviance Chisq Df Pr(>Chisq)
## m1 3 37.478 46.127 -15.739 31.479
## m2 5 31.720 46.134 -10.860 21.720 9.7583 2 0.007603 **
## m3 6 33.210 50.507 -10.605 21.210 0.5101 1 0.475083
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1tab_model(m1,m2,m3)| meanPerf.x | meanPerf.x | meanPerf.x | |||||||
|---|---|---|---|---|---|---|---|---|---|
| Predictors | Estimates | CI | p | Estimates | CI | p | Estimates | CI | p |
| (Intercept) | 0.56 | 0.50 – 0.61 | <0.001 | 0.23 | -0.06 – 0.51 | 0.115 | 0.14 | -0.22 – 0.51 | 0.431 |
| hierarchy [H2] | -0.11 | -0.19 – -0.03 | 0.008 | 0.07 | -0.44 – 0.58 | 0.781 | |||
| meanPerf y | 0.49 | 0.13 – 0.86 | 0.008 | 0.60 | 0.13 – 1.07 | 0.013 | |||
|
hierarchy [H2] × meanPerf y |
-0.23 | -0.86 – 0.41 | 0.483 | ||||||
| Random Effects | |||||||||
| σ2 | 0.06 | 0.05 | 0.05 | ||||||
| τ00 | 0.02 participant | 0.03 participant | 0.03 participant | ||||||
| ICC | 0.28 | 0.40 | 0.40 | ||||||
| N | 71 participant | 71 participant | 71 participant | ||||||
| Observations | 132 | 132 | 132 | ||||||
| Marginal R2 / Conditional R2 | 0.000 / 0.280 | 0.068 / 0.442 | 0.070 / 0.447 | ||||||
plot_model(m2, type="pred", terms = c("meanPerf.y","hierarchy"))
Looking into Distance effects
- no linear effect of distance
load("awareness_score_scenes.Rdata")
load("awareness_score_facescene.Rdata")
load("awareness_score_shape.Rdata")
load("awareness_score_obj.Rdata")
load("session3_pretty_filtered_obj.Rdata")
lmm.scenes <- bind_rows(session3.pretty.filtered.scenes %>% group_by(participant,pairType, awareness) %>% summarise(H1 = mean(H1), H2 = mean(H2)) %>% gather(hierarchy, meanPerf, c("H1","H2")) %>% mutate(session=3) ,rbind(session1.avgs.scenes, session2.avgs.scenes) %>% filter(participant %in% session3.pretty.filtered.scenes$participant) %>% group_by(participant,session,blockcount) %>% summarise(premise=mean(key_resp_im.corr)) %>% gather(pairType, meanPerf, premise) %>% mutate(hierarchy=paste0("H",session))) %>% mutate(stimuli="scene") %>% select(-awareness) %>% left_join(awareness_score.scenes, c("participant"="code")) %>% mutate(participant=paste0("scn",participant), timeelapsed = timeelapsed*60)## `summarise()` has grouped output by 'participant', 'pairType'. You can override
## using the `.groups` argument.
## `summarise()` has grouped output by 'participant', 'session'. You can override
## using the `.groups` argument.lmm.facescene <- bind_rows(session3.pretty.filtered.facescene %>% group_by(participant,pairType, awareness) %>% summarise(H1 = mean(H1), H2 = mean(H2)) %>% gather(hierarchy, meanPerf, c("H1","H2")) %>% mutate(session=3), rbind(session1.avgs.facescene, session2.avgs.facescene) %>% filter(participant %in% session3.pretty.filtered.facescene$participant) %>% group_by(participant,session,blockcount) %>% summarise(premise=mean(key_resp_im.corr)) %>% gather(pairType, meanPerf, premise) %>% mutate(hierarchy=paste0("H",session))) %>% mutate(stimuli="facescene") %>% select(-awareness) %>% left_join(awareness_score.facescene, c("participant"="code")) %>% mutate(participant=paste0("fsc",participant))## `summarise()` has grouped output by 'participant', 'pairType'. You can override
## using the `.groups` argument.
## `summarise()` has grouped output by 'participant', 'session'. You can override
## using the `.groups` argument.lmm.shape <- bind_rows(session3.pretty.filtered.shape %>% group_by(participant,pairType, awareness) %>% summarise(H1 = mean(H1), H2 = mean(H2)) %>% gather(hierarchy, meanPerf, c("H1","H2")) %>% mutate(session=3), rbind(session1.avgs.shape, session2.avgs.shape) %>% filter(participant %in% session3.pretty.filtered.shape$participant) %>% group_by(participant,session,blockcount) %>% summarise(premise=mean(key_resp_im.corr)) %>% gather(pairType, meanPerf, premise) %>% mutate(hierarchy=paste0("H",session))) %>% mutate(stimuli="obj") %>% select(-awareness) %>% left_join(awareness_score.shape, c("participant"="code")) %>% mutate(participant=paste0("o_s",participant))## `summarise()` has grouped output by 'participant', 'pairType'. You can override
## using the `.groups` argument.
## `summarise()` has grouped output by 'participant', 'session'. You can override
## using the `.groups` argument.# save(session3.pretty.filtered.obj, file = "session3_pretty_filtered_obj.Rdata")
#load("session3_pretty_filtered_obj.Rdata")
lmm.noshape <- bind_rows(session3.pretty.filtered.obj %>% group_by(participant,pairType, awareness) %>% summarise(H1 = mean(H1), H2 = mean(H2)) %>% gather(hierarchy, meanPerf, c("H1","H2")) %>% mutate(session=3), rbind(session1.avgs.obj, session2.avgs.obj) %>% filter(participant %in% session3.pretty.filtered.obj$participant) %>% group_by(participant,session,blockcount) %>% summarise(premise=mean(key_resp_im.corr)) %>% gather(pairType, meanPerf, premise) %>% mutate(hierarchy=paste0("H",session))) %>% mutate(stimuli="obj") %>% select(-awareness) %>% left_join(awareness_score.obj, c("participant"="code")) %>% mutate(participant=paste0("o_ns",participant))## `summarise()` has grouped output by 'participant', 'pairType'. You can override
## using the `.groups` argument.
## `summarise()` has grouped output by 'participant', 'session'. You can override
## using the `.groups` argument.lmm.combined <- bind_rows(lmm.scenes,lmm.facescene,lmm.shape,lmm.noshape)
s1s2.im <- lmm.combined %>% filter(session %in% c(1,2))
lmm.combined.wide <- lmm.combined %>% filter(session == 3) %>% left_join(.,s1s2.im,by=c("participant"="participant","hierarchy"="hierarchy")) %>% filter(pairType.x != "premise") %>%
mutate(Rank1 = as.numeric(str_sub(pairType.x, start = 1, end = 1)),
Rank2 = as.numeric(str_sub(pairType.x, start = 2, end = 2))) %>% mutate(distance = abs(Rank1-Rank2), totalrank=Rank1+Rank2)
glimpse(lmm.combined.wide)## Rows: 584
## Columns: 20
## Groups: participant [73]
## $ participant <chr> "scnBIHKJQY", "scnBIHKJQY", "scnBIHKJQY", "scnBIHKJQY", …
## $ pairType.x <chr> "16", "24", "25", "35", "16", "24", "25", "35", "16", "2…
## $ hierarchy <chr> "H1", "H1", "H1", "H1", "H1", "H1", "H1", "H1", "H1", "H…
## $ meanPerf.x <dbl> 0.875, 0.125, 0.250, 0.000, 0.625, 0.000, 0.125, 0.875, …
## $ session.x <dbl> 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3,…
## $ blockcount.x <dbl> NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, …
## $ stimuli.x <chr> "scene", "scene", "scene", "scene", "scene", "scene", "s…
## $ awareness.x <dbl> 8, 8, 8, 8, 8, 8, 8, 8, 3, 3, 3, 3, 5, 5, 5, 5, 11, 11, …
## $ timeelapsed.x <dbl> 186.9333, 186.9333, 186.9333, 186.9333, 204.5167, 204.51…
## $ pairType.y <chr> "premise", "premise", "premise", "premise", "premise", "…
## $ meanPerf.y <dbl> 0.5500, 0.5500, 0.5500, 0.5500, 0.5500, 0.5500, 0.5500, …
## $ session.y <dbl> 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,…
## $ blockcount.y <dbl> 9, 9, 9, 9, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 8, 8, 8,…
## $ stimuli.y <chr> "scene", "scene", "scene", "scene", "scene", "scene", "s…
## $ awareness.y <dbl> 8, 8, 8, 8, 8, 8, 8, 8, 3, 3, 3, 3, 5, 5, 5, 5, 11, 11, …
## $ timeelapsed.y <dbl> 186.9333, 186.9333, 186.9333, 186.9333, 204.5167, 204.51…
## $ Rank1 <dbl> 1, 2, 2, 3, 1, 2, 2, 3, 1, 2, 2, 3, 1, 2, 2, 3, 1, 2, 2,…
## $ Rank2 <dbl> 6, 4, 5, 5, 6, 4, 5, 5, 6, 4, 5, 5, 6, 4, 5, 5, 6, 4, 5,…
## $ distance <dbl> 5, 2, 3, 2, 5, 2, 3, 2, 5, 2, 3, 2, 5, 2, 3, 2, 5, 2, 3,…
## $ totalrank <dbl> 7, 6, 7, 8, 7, 6, 7, 8, 7, 6, 7, 8, 7, 6, 7, 8, 7, 6, 7,…library(lme4)
library(lmerTest)
# new attempt : with wide
# IMPORTANT
m0 <- lmer(meanPerf.x ~ 1 + (1 | participant), data = lmm.combined.wide %>% filter(pairType.x != "16"))
m1 <- lmer(meanPerf.x ~ hierarchy + (1 | participant), data = lmm.combined.wide %>% filter(pairType.x != "16"))
m2 <- lmer(meanPerf.x ~ hierarchy + meanPerf.y + (1 | participant), data = lmm.combined.wide %>% filter(pairType.x != "16"))
m3 <- lmer(meanPerf.x ~ hierarchy * meanPerf.y + (1 | participant) , data = lmm.combined.wide %>% filter(pairType.x != "16"))
tab_model(m0,m1,m2,m3)| meanPerf.x | meanPerf.x | meanPerf.x | meanPerf.x | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Predictors | Estimates | CI | p | Estimates | CI | p | Estimates | CI | p | Estimates | CI | p |
| (Intercept) | 0.54 | 0.49 – 0.59 | <0.001 | 0.56 | 0.50 – 0.62 | <0.001 | 0.11 | -0.09 – 0.31 | 0.265 | -0.04 | -0.27 – 0.19 | 0.726 |
| hierarchy [H2] | -0.04 | -0.10 – 0.02 | 0.233 | -0.09 | -0.15 – -0.03 | 0.004 | 0.30 | -0.03 – 0.62 | 0.072 | |||
| meanPerf y | 0.62 | 0.36 – 0.89 | <0.001 | 0.84 | 0.53 – 1.15 | <0.001 | ||||||
|
hierarchy [H2] × meanPerf y |
-0.50 | -0.91 – -0.09 | 0.017 | |||||||||
| Random Effects | ||||||||||||
| σ2 | 0.10 | 0.10 | 0.09 | 0.09 | ||||||||
| τ00 | 0.03 participant | 0.03 participant | 0.03 participant | 0.04 participant | ||||||||
| ICC | 0.23 | 0.23 | 0.27 | 0.28 | ||||||||
| N | 73 participant | 73 participant | 73 participant | 73 participant | ||||||||
| Observations | 438 | 438 | 438 | 438 | ||||||||
| Marginal R2 / Conditional R2 | 0.000 / 0.226 | 0.003 / 0.228 | 0.070 / 0.322 | 0.080 / 0.336 | ||||||||
m4 <- lmer(meanPerf.x ~ hierarchy + meanPerf.y + distance + (1 | participant) , data = lmm.combined.wide %>% filter(pairType.x != "16"))
m5 <- lmer(meanPerf.x ~ hierarchy * meanPerf.y + distance + (1 | participant) , data = lmm.combined.wide %>% filter(pairType.x != "16"))
m6 <- lmer(meanPerf.x ~ hierarchy * meanPerf.y + distance*hierarchy + (1 | participant) , data = lmm.combined.wide %>% filter(pairType.x != "16"))
tab_model(m1,m2,m3, m4,m5,m6)| meanPerf.x | meanPerf.x | meanPerf.x | meanPerf.x | meanPerf.x | meanPerf.x | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Predictors | Estimates | CI | p | Estimates | CI | p | Estimates | CI | p | Estimates | CI | p | Estimates | CI | p | Estimates | CI | p |
| (Intercept) | 0.56 | 0.50 – 0.62 | <0.001 | 0.11 | -0.09 – 0.31 | 0.265 | -0.04 | -0.27 – 0.19 | 0.726 | 0.01 | -0.24 – 0.25 | 0.956 | -0.15 | -0.42 – 0.12 | 0.287 | -0.19 | -0.50 – 0.11 | 0.214 |
| hierarchy [H2] | -0.04 | -0.10 – 0.02 | 0.233 | -0.09 | -0.15 – -0.03 | 0.004 | 0.30 | -0.03 – 0.62 | 0.072 | -0.09 | -0.15 – -0.03 | 0.004 | 0.30 | -0.03 – 0.62 | 0.071 | 0.39 | -0.04 – 0.82 | 0.074 |
| meanPerf y | 0.62 | 0.36 – 0.89 | <0.001 | 0.84 | 0.53 – 1.15 | <0.001 | 0.63 | 0.36 – 0.89 | <0.001 | 0.84 | 0.53 – 1.15 | <0.001 | 0.84 | 0.53 – 1.15 | <0.001 | |||
|
hierarchy [H2] × meanPerf y |
-0.50 | -0.91 – -0.09 | 0.017 | -0.50 | -0.91 – -0.09 | 0.016 | -0.50 | -0.91 – -0.09 | 0.017 | |||||||||
| distance | 0.05 | -0.02 – 0.11 | 0.144 | 0.05 | -0.01 – 0.11 | 0.140 | 0.07 | -0.02 – 0.15 | 0.133 | |||||||||
| hierarchy [H2] × distance | -0.04 | -0.16 – 0.08 | 0.514 | |||||||||||||||
| Random Effects | ||||||||||||||||||
| σ2 | 0.10 | 0.09 | 0.09 | 0.09 | 0.09 | 0.09 | ||||||||||||
| τ00 | 0.03 participant | 0.03 participant | 0.04 participant | 0.03 participant | 0.04 participant | 0.04 participant | ||||||||||||
| ICC | 0.23 | 0.27 | 0.28 | 0.27 | 0.28 | 0.28 | ||||||||||||
| N | 73 participant | 73 participant | 73 participant | 73 participant | 73 participant | 73 participant | ||||||||||||
| Observations | 438 | 438 | 438 | 438 | 438 | 438 | ||||||||||||
| Marginal R2 / Conditional R2 | 0.003 / 0.228 | 0.070 / 0.322 | 0.080 / 0.336 | 0.074 / 0.325 | 0.083 / 0.339 | 0.084 / 0.339 | ||||||||||||
# distance doesn't improve LogLik
anova(m1,m2,m3,m4)## refitting model(s) with ML (instead of REML)## Data: lmm.combined.wide %>% filter(pairType.x != "16")
## Models:
## m1: meanPerf.x ~ hierarchy + (1 | participant)
## m2: meanPerf.x ~ hierarchy + meanPerf.y + (1 | participant)
## m3: meanPerf.x ~ hierarchy * meanPerf.y + (1 | participant)
## m4: meanPerf.x ~ hierarchy + meanPerf.y + distance + (1 | participant)
## npar AIC BIC logLik deviance Chisq Df Pr(>Chisq)
## m1 4 310.99 327.32 -151.50 302.99
## m2 5 292.57 312.98 -141.28 282.57 20.4228 1 6.208e-06 ***
## m3 6 288.82 313.31 -138.41 276.82 5.7452 1 0.01653 *
## m4 6 292.41 316.90 -140.21 280.41 0.0000 0
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1plot_model(m3, type="pred", terms = c("meanPerf.y","hierarchy"))
plot_model(m3, type="diag")## Warning in checkMatrixPackageVersion(): Package version inconsistency detected.
## TMB was built with Matrix version 1.5.3
## Current Matrix version is 1.5.1
## Please re-install 'TMB' from source using install.packages('TMB', type = 'source') or ask CRAN for a binary version of 'TMB' matching CRAN's 'Matrix' package## [[1]]## `geom_smooth()` using formula = 'y ~ x'
##
## [[2]]
## [[2]]$participant## `geom_smooth()` using formula = 'y ~ x'
##
##
## [[3]]
##
## [[4]]## `geom_smooth()` using formula = 'y ~ x'
# new attempt : with wide; without below chance immediate premise performers // affects two participants
m1 <- lmer(meanPerf.x ~ 1 + (1 | participant), data = lmm.combined.wide %>% filter(pairType.x != "16", meanPerf.y>0.5), REML=FALSE)
m2 <- lmer(meanPerf.x ~ hierarchy + meanPerf.y + (1 | participant), data = lmm.combined.wide %>% filter(pairType.x != "16",meanPerf.y>0.5))
m3 <- lmer(meanPerf.x ~ hierarchy * meanPerf.y + (1 | participant) , data = lmm.combined.wide %>% filter(pairType.x != "16",meanPerf.y>0.5))
m4 <- lmer(meanPerf.x ~ hierarchy + meanPerf.y + distance + (1 | participant) , data = lmm.combined.wide %>% filter(pairType.x != "16",meanPerf.y>0.5))
m5 <- lmer(meanPerf.x ~ hierarchy * meanPerf.y + distance + (1 | participant) , data = lmm.combined.wide %>% filter(pairType.x != "16",meanPerf.y>0.5))
m6 <- lmer(meanPerf.x ~ hierarchy * meanPerf.y + distance*hierarchy + (1 | participant) , data = lmm.combined.wide %>% filter(pairType.x != "16",meanPerf.y>0.5))
tab_model(m1,m2,m3,m4,m5,m6)| meanPerf.x | meanPerf.x | meanPerf.x | meanPerf.x | meanPerf.x | meanPerf.x | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Predictors | Estimates | CI | p | Estimates | CI | p | Estimates | CI | p | Estimates | CI | p | Estimates | CI | p | Estimates | CI | p |
| (Intercept) | 0.56 | 0.51 – 0.61 | <0.001 | 0.15 | -0.10 – 0.40 | 0.245 | 0.06 | -0.26 – 0.37 | 0.720 | 0.06 | -0.23 – 0.35 | 0.695 | -0.03 | -0.38 – 0.31 | 0.849 | -0.07 | -0.45 – 0.31 | 0.708 |
| hierarchy [H2] | -0.12 | -0.18 – -0.05 | <0.001 | 0.09 | -0.34 – 0.52 | 0.686 | -0.12 | -0.18 – -0.05 | <0.001 | 0.09 | -0.34 – 0.52 | 0.686 | 0.16 | -0.36 – 0.68 | 0.538 | |||
| meanPerf y | 0.60 | 0.28 – 0.92 | <0.001 | 0.72 | 0.31 – 1.12 | 0.001 | 0.60 | 0.28 – 0.92 | <0.001 | 0.72 | 0.31 – 1.12 | 0.001 | 0.72 | 0.31 – 1.12 | 0.001 | |||
|
hierarchy [H2] × meanPerf y |
-0.26 | -0.79 – 0.27 | 0.342 | -0.26 | -0.79 – 0.27 | 0.341 | -0.26 | -0.79 – 0.27 | 0.342 | |||||||||
| distance | 0.04 | -0.02 – 0.10 | 0.224 | 0.04 | -0.02 – 0.10 | 0.224 | 0.06 | -0.04 – 0.15 | 0.230 | |||||||||
| hierarchy [H2] × distance | -0.03 | -0.16 – 0.09 | 0.617 | |||||||||||||||
| Random Effects | ||||||||||||||||||
| σ2 | 0.10 | 0.09 | 0.09 | 0.09 | 0.09 | 0.09 | ||||||||||||
| τ00 | 0.03 participant | 0.04 participant | 0.04 participant | 0.04 participant | 0.04 participant | 0.04 participant | ||||||||||||
| ICC | 0.26 | 0.31 | 0.31 | 0.31 | 0.31 | 0.31 | ||||||||||||
| N | 71 participant | 71 participant | 71 participant | 71 participant | 71 participant | 71 participant | ||||||||||||
| Observations | 396 | 396 | 396 | 396 | 396 | 396 | ||||||||||||
| Marginal R2 / Conditional R2 | 0.000 / 0.261 | 0.055 / 0.350 | 0.057 / 0.353 | 0.058 / 0.352 | 0.059 / 0.355 | 0.059 / 0.354 | ||||||||||||
Some descriptive plots
lmm.combined.wide %>% group_by(participant, hierarchy) %>% count()## # A tibble: 146 × 3
## # Groups: participant, hierarchy [146]
## participant hierarchy n
## <chr> <chr> <int>
## 1 fscBTCQUV H1 4
## 2 fscBTCQUV H2 4
## 3 fscCERPNO H1 4
## 4 fscCERPNO H2 4
## 5 fscFOMAYQ H1 4
## 6 fscFOMAYQ H2 4
## 7 fscFTUDJP H1 4
## 8 fscFTUDJP H2 4
## 9 fscGLIBTX H1 4
## 10 fscGLIBTX H2 4
## # … with 136 more rowslmm.combined.wide %>% filter(pairType.x != "16", meanPerf.y>0.5) %>% group_by(participant, hierarchy) %>% count()## # A tibble: 132 × 3
## # Groups: participant, hierarchy [132]
## participant hierarchy n
## <chr> <chr> <int>
## 1 fscBTCQUV H1 3
## 2 fscBTCQUV H2 3
## 3 fscCERPNO H1 3
## 4 fscCERPNO H2 3
## 5 fscFOMAYQ H1 3
## 6 fscFOMAYQ H2 3
## 7 fscFTUDJP H1 3
## 8 fscFTUDJP H2 3
## 9 fscGLIBTX H1 3
## 10 fscGLIBTX H2 3
## # … with 122 more rowslibrary(ggstatsplot)
tmp <- lmm.combined.wide %>% filter(pairType.x != "16") %>% filter(meanPerf.y>0.5)
grouped_ggwithinstats(
data = lmm.combined.wide %>% filter(pairType.x != "16", participant %in% tmp$participant),
x = hierarchy,
y = meanPerf.x,
grouping.var = pairType.x ,
type = "np",
bf.message = FALSE
)
# IMPORTANT
# distant inference pair
m0 <- lmer(meanPerf.x ~ 1 + (1 | participant), data = lmm.combined.wide %>% filter(pairType.x == "25"))
m1 <- lmer(meanPerf.x ~ hierarchy + (1 | participant), data = lmm.combined.wide %>% filter(pairType.x == "25"))
m2 <- lmer(meanPerf.x ~ hierarchy + meanPerf.y + (1 | participant), data = lmm.combined.wide %>% filter(pairType.x == "25"))
m3 <- lmer(meanPerf.x ~ hierarchy * meanPerf.y + (1 | participant) , data = lmm.combined.wide %>% filter(pairType.x == "25"))
tab_model(m0,m1,m2,m3)| meanPerf.x | meanPerf.x | meanPerf.x | meanPerf.x | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Predictors | Estimates | CI | p | Estimates | CI | p | Estimates | CI | p | Estimates | CI | p |
| (Intercept) | 0.57 | 0.51 – 0.63 | <0.001 | 0.60 | 0.52 – 0.68 | <0.001 | 0.23 | -0.04 – 0.51 | 0.096 | 0.09 | -0.25 – 0.42 | 0.614 |
| hierarchy [H2] | -0.06 | -0.16 – 0.04 | 0.208 | -0.11 | -0.21 – -0.01 | 0.032 | 0.25 | -0.25 – 0.75 | 0.324 | |||
| meanPerf y | 0.52 | 0.15 – 0.88 | 0.007 | 0.72 | 0.26 – 1.18 | 0.002 | ||||||
|
hierarchy [H2] × meanPerf y |
-0.47 | -1.11 – 0.17 | 0.151 | |||||||||
| Random Effects | ||||||||||||
| σ2 | 0.09 | 0.09 | 0.08 | 0.08 | ||||||||
| τ00 | 0.03 participant | 0.03 participant | 0.03 participant | 0.04 participant | ||||||||
| ICC | 0.23 | 0.24 | 0.29 | 0.31 | ||||||||
| N | 73 participant | 73 participant | 73 participant | 73 participant | ||||||||
| Observations | 146 | 146 | 146 | 146 | ||||||||
| Marginal R2 / Conditional R2 | 0.000 / 0.233 | 0.008 / 0.243 | 0.061 / 0.336 | 0.071 / 0.359 | ||||||||
plot_model(m2, type="pred", terms = c("meanPerf.y","hierarchy"))
plot_model(m2, type="diag")## [[1]]## `geom_smooth()` using formula = 'y ~ x'
##
## [[2]]
## [[2]]$participant## `geom_smooth()` using formula = 'y ~ x'
##
##
## [[3]]
##
## [[4]]## `geom_smooth()` using formula = 'y ~ x'
# BD inference pair
m0 <- lmer(meanPerf.x ~ 1 + (1 | participant), data = lmm.combined.wide %>% filter(pairType.x == "24"))
m1 <- lmer(meanPerf.x ~ hierarchy + (1 | participant), data = lmm.combined.wide %>% filter(pairType.x == "24"))
m2 <- lmer(meanPerf.x ~ hierarchy + meanPerf.y + (1 | participant), data = lmm.combined.wide %>% filter(pairType.x == "24"))
m3 <- lmer(meanPerf.x ~ hierarchy * meanPerf.y + (1 | participant) , data = lmm.combined.wide %>% filter(pairType.x == "24"))
tab_model(m0,m1,m2,m3)| meanPerf.x | meanPerf.x | meanPerf.x | meanPerf.x | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Predictors | Estimates | CI | p | Estimates | CI | p | Estimates | CI | p | Estimates | CI | p |
| (Intercept) | 0.52 | 0.46 – 0.59 | <0.001 | 0.55 | 0.47 – 0.64 | <0.001 | 0.35 | 0.05 – 0.65 | 0.024 | 0.14 | -0.25 – 0.52 | 0.476 |
| hierarchy [H2] | -0.06 | -0.18 – 0.06 | 0.326 | -0.09 | -0.21 – 0.04 | 0.172 | 0.41 | -0.20 – 1.01 | 0.184 | |||
| meanPerf y | 0.29 | -0.11 – 0.69 | 0.154 | 0.58 | 0.06 – 1.10 | 0.030 | ||||||
|
hierarchy [H2] × meanPerf y |
-0.64 | -1.42 – 0.13 | 0.102 | |||||||||
| Random Effects | ||||||||||||
| σ2 | 0.13 | 0.13 | 0.13 | 0.12 | ||||||||
| τ00 | 0.00 participant | 0.00 participant | 0.01 participant | 0.01 participant | ||||||||
| ICC | 0.04 | 0.04 | 0.07 | 0.10 | ||||||||
| N | 73 participant | 73 participant | 73 participant | 73 participant | ||||||||
| Observations | 146 | 146 | 146 | 146 | ||||||||
| Marginal R2 / Conditional R2 | 0.000 / 0.036 | 0.006 / 0.042 | 0.021 / 0.086 | 0.038 / 0.131 | ||||||||
grouped_ggbetweenstats(
data = lmm.combined.wide %>% filter(pairType.x != "16"),
x = pairType.x,
y = meanPerf.x,
grouping.var = hierarchy ,
type = "p",
bf.message = FALSE
)
Looking into Total rank
# new attempt : with wide
m0 <- lmer(meanPerf.x ~ 1 + (1 | participant), data = lmm.combined.wide %>% filter(pairType.x != "16"))
m1 <- lmer(meanPerf.x ~ hierarchy + (1 | participant), data = lmm.combined.wide %>% filter(pairType.x != "16"))
m2 <- lmer(meanPerf.x ~ hierarchy + meanPerf.y + (1 | participant), data = lmm.combined.wide %>% filter(pairType.x != "16"))
m3 <- lmer(meanPerf.x ~ hierarchy * meanPerf.y + (1 | participant) , data = lmm.combined.wide %>% filter(pairType.x != "16"))
tab_model(m0,m1,m2,m3)| meanPerf.x | meanPerf.x | meanPerf.x | meanPerf.x | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Predictors | Estimates | CI | p | Estimates | CI | p | Estimates | CI | p | Estimates | CI | p |
| (Intercept) | 0.54 | 0.49 – 0.59 | <0.001 | 0.56 | 0.50 – 0.62 | <0.001 | 0.11 | -0.09 – 0.31 | 0.265 | -0.04 | -0.27 – 0.19 | 0.726 |
| hierarchy [H2] | -0.04 | -0.10 – 0.02 | 0.233 | -0.09 | -0.15 – -0.03 | 0.004 | 0.30 | -0.03 – 0.62 | 0.072 | |||
| meanPerf y | 0.62 | 0.36 – 0.89 | <0.001 | 0.84 | 0.53 – 1.15 | <0.001 | ||||||
|
hierarchy [H2] × meanPerf y |
-0.50 | -0.91 – -0.09 | 0.017 | |||||||||
| Random Effects | ||||||||||||
| σ2 | 0.10 | 0.10 | 0.09 | 0.09 | ||||||||
| τ00 | 0.03 participant | 0.03 participant | 0.03 participant | 0.04 participant | ||||||||
| ICC | 0.23 | 0.23 | 0.27 | 0.28 | ||||||||
| N | 73 participant | 73 participant | 73 participant | 73 participant | ||||||||
| Observations | 438 | 438 | 438 | 438 | ||||||||
| Marginal R2 / Conditional R2 | 0.000 / 0.226 | 0.003 / 0.228 | 0.070 / 0.322 | 0.080 / 0.336 | ||||||||
m4 <- lmer(meanPerf.x ~ hierarchy + meanPerf.y + totalrank + (1 | participant) , data = lmm.combined.wide %>% filter(pairType.x != "16"))
m5 <- lmer(meanPerf.x ~ hierarchy * meanPerf.y + totalrank + (1 | participant) , data = lmm.combined.wide %>% filter(pairType.x != "16"))
m6 <- lmer(meanPerf.x ~ hierarchy * meanPerf.y + totalrank*hierarchy + (1 | participant) , data = lmm.combined.wide %>% filter(pairType.x != "16"))
tab_model(m1,m2,m3, m4,m5,m6)| meanPerf.x | meanPerf.x | meanPerf.x | meanPerf.x | meanPerf.x | meanPerf.x | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Predictors | Estimates | CI | p | Estimates | CI | p | Estimates | CI | p | Estimates | CI | p | Estimates | CI | p | Estimates | CI | p |
| (Intercept) | 0.56 | 0.50 – 0.62 | <0.001 | 0.11 | -0.09 – 0.31 | 0.265 | -0.04 | -0.27 – 0.19 | 0.726 | 0.10 | -0.22 – 0.41 | 0.554 | -0.06 | -0.40 – 0.28 | 0.731 | 0.07 | -0.34 – 0.49 | 0.738 |
| hierarchy [H2] | -0.04 | -0.10 – 0.02 | 0.233 | -0.09 | -0.15 – -0.03 | 0.004 | 0.30 | -0.03 – 0.62 | 0.072 | -0.09 | -0.15 – -0.03 | 0.004 | 0.30 | -0.03 – 0.62 | 0.072 | 0.04 | -0.55 – 0.62 | 0.904 |
| meanPerf y | 0.62 | 0.36 – 0.89 | <0.001 | 0.84 | 0.53 – 1.15 | <0.001 | 0.62 | 0.36 – 0.89 | <0.001 | 0.84 | 0.52 – 1.15 | <0.001 | 0.84 | 0.52 – 1.15 | <0.001 | |||
|
hierarchy [H2] × meanPerf y |
-0.50 | -0.91 – -0.09 | 0.017 | -0.50 | -0.91 – -0.09 | 0.017 | -0.50 | -0.91 – -0.09 | 0.017 | |||||||||
| totalrank | 0.00 | -0.03 – 0.04 | 0.885 | 0.00 | -0.03 – 0.04 | 0.885 | -0.02 | -0.07 – 0.03 | 0.523 | |||||||||
|
hierarchy [H2] × totalrank |
0.04 | -0.03 – 0.11 | 0.295 | |||||||||||||||
| Random Effects | ||||||||||||||||||
| σ2 | 0.10 | 0.09 | 0.09 | 0.09 | 0.09 | 0.09 | ||||||||||||
| τ00 | 0.03 participant | 0.03 participant | 0.04 participant | 0.03 participant | 0.04 participant | 0.04 participant | ||||||||||||
| ICC | 0.23 | 0.27 | 0.28 | 0.27 | 0.28 | 0.28 | ||||||||||||
| N | 73 participant | 73 participant | 73 participant | 73 participant | 73 participant | 73 participant | ||||||||||||
| Observations | 438 | 438 | 438 | 438 | 438 | 438 | ||||||||||||
| Marginal R2 / Conditional R2 | 0.003 / 0.228 | 0.070 / 0.322 | 0.080 / 0.336 | 0.070 / 0.321 | 0.080 / 0.335 | 0.081 / 0.336 | ||||||||||||
# distance doesn't improve LogLik
anova(m1,m2,m3,m4)## refitting model(s) with ML (instead of REML)## Data: lmm.combined.wide %>% filter(pairType.x != "16")
## Models:
## m1: meanPerf.x ~ hierarchy + (1 | participant)
## m2: meanPerf.x ~ hierarchy + meanPerf.y + (1 | participant)
## m3: meanPerf.x ~ hierarchy * meanPerf.y + (1 | participant)
## m4: meanPerf.x ~ hierarchy + meanPerf.y + totalrank + (1 | participant)
## npar AIC BIC logLik deviance Chisq Df Pr(>Chisq)
## m1 4 310.99 327.32 -151.50 302.99
## m2 5 292.57 312.98 -141.28 282.57 20.4228 1 6.208e-06 ***
## m3 6 288.82 313.31 -138.41 276.82 5.7452 1 0.01653 *
## m4 6 294.55 319.04 -141.27 282.55 0.0000 0
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1plot_model(m3, type="pred", terms = c("meanPerf.y","hierarchy"))
# new attempt : with wide; without below chance immediate premise performers // affects two participants
m1 <- lmer(meanPerf.x ~ 1 + (1 | participant), data = lmm.combined.wide %>% filter(pairType.x != "16", meanPerf.y>0.5), REML=FALSE)
m2 <- lmer(meanPerf.x ~ hierarchy + meanPerf.y + (1 | participant), data = lmm.combined.wide %>% filter(pairType.x != "16",meanPerf.y>0.5))
m3 <- lmer(meanPerf.x ~ hierarchy * meanPerf.y + (1 | participant) , data = lmm.combined.wide %>% filter(pairType.x != "16",meanPerf.y>0.5))
m4 <- lmer(meanPerf.x ~ hierarchy + meanPerf.y + distance + (1 | participant) , data = lmm.combined.wide %>% filter(pairType.x != "16",meanPerf.y>0.5))
m5 <- lmer(meanPerf.x ~ hierarchy * meanPerf.y + distance + (1 | participant) , data = lmm.combined.wide %>% filter(pairType.x != "16",meanPerf.y>0.5))
m6 <- lmer(meanPerf.x ~ hierarchy * meanPerf.y + distance*hierarchy + (1 | participant) , data = lmm.combined.wide %>% filter(pairType.x != "16",meanPerf.y>0.5))
tab_model(m1,m2,m3,m4,m5,m6)| meanPerf.x | meanPerf.x | meanPerf.x | meanPerf.x | meanPerf.x | meanPerf.x | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Predictors | Estimates | CI | p | Estimates | CI | p | Estimates | CI | p | Estimates | CI | p | Estimates | CI | p | Estimates | CI | p |
| (Intercept) | 0.56 | 0.51 – 0.61 | <0.001 | 0.15 | -0.10 – 0.40 | 0.245 | 0.06 | -0.26 – 0.37 | 0.720 | 0.06 | -0.23 – 0.35 | 0.695 | -0.03 | -0.38 – 0.31 | 0.849 | -0.07 | -0.45 – 0.31 | 0.708 |
| hierarchy [H2] | -0.12 | -0.18 – -0.05 | <0.001 | 0.09 | -0.34 – 0.52 | 0.686 | -0.12 | -0.18 – -0.05 | <0.001 | 0.09 | -0.34 – 0.52 | 0.686 | 0.16 | -0.36 – 0.68 | 0.538 | |||
| meanPerf y | 0.60 | 0.28 – 0.92 | <0.001 | 0.72 | 0.31 – 1.12 | 0.001 | 0.60 | 0.28 – 0.92 | <0.001 | 0.72 | 0.31 – 1.12 | 0.001 | 0.72 | 0.31 – 1.12 | 0.001 | |||
|
hierarchy [H2] × meanPerf y |
-0.26 | -0.79 – 0.27 | 0.342 | -0.26 | -0.79 – 0.27 | 0.341 | -0.26 | -0.79 – 0.27 | 0.342 | |||||||||
| distance | 0.04 | -0.02 – 0.10 | 0.224 | 0.04 | -0.02 – 0.10 | 0.224 | 0.06 | -0.04 – 0.15 | 0.230 | |||||||||
| hierarchy [H2] × distance | -0.03 | -0.16 – 0.09 | 0.617 | |||||||||||||||
| Random Effects | ||||||||||||||||||
| σ2 | 0.10 | 0.09 | 0.09 | 0.09 | 0.09 | 0.09 | ||||||||||||
| τ00 | 0.03 participant | 0.04 participant | 0.04 participant | 0.04 participant | 0.04 participant | 0.04 participant | ||||||||||||
| ICC | 0.26 | 0.31 | 0.31 | 0.31 | 0.31 | 0.31 | ||||||||||||
| N | 71 participant | 71 participant | 71 participant | 71 participant | 71 participant | 71 participant | ||||||||||||
| Observations | 396 | 396 | 396 | 396 | 396 | 396 | ||||||||||||
| Marginal R2 / Conditional R2 | 0.000 / 0.261 | 0.055 / 0.350 | 0.057 / 0.353 | 0.058 / 0.352 | 0.059 / 0.355 | 0.059 / 0.354 | ||||||||||||