Analysis: Is habitat selection in the wild shaped by individual-level cognitive biases in orientation strategy?
Beardsworth et al. 2021
Import Libraries
library(MuMIn)
library(ggplot2)
library(lme4)
library(extrafont)
library(ggbeeswarm)
options(scipen=999)Cognition Analysis
learning <- read.csv("mazeData.csv")
# Did pheasants learn the maze task (errors)
m1 <- glmer(NumErrors ~ TrialNumber + Sex + Group + (1|Bird), data=learning, family = poisson(link=log), na.action="na.fail")
summary(m1)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: poisson ( log )
## Formula: NumErrors ~ TrialNumber + Sex + Group + (1 | Bird)
## Data: learning
##
## AIC BIC logLik deviance df.resid
## 4603.9 4626.1 -2297.0 4593.9 619
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -3.8277 -1.5327 -0.5791 0.6057 13.1994
##
## Random effects:
## Groups Name Variance Std.Dev.
## Bird (Intercept) 0.2944 0.5426
## Number of obs: 624, groups: Bird, 78
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 1.484583 0.111783 13.281 < 0.0000000000000002 ***
## TrialNumber -0.050948 0.008641 -5.896 0.00000000372 ***
## SexM -0.061897 0.130985 -0.473 0.637
## GroupExperimental 0.058010 0.131821 0.440 0.660
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) TrlNmb SexM
## TrialNumber -0.327
## SexM -0.531 0.000
## GrpExprmntl -0.491 0.000 -0.048output <- dredge(m1)
output## Global model call: glmer(formula = NumErrors ~ TrialNumber + Sex + Group + (1 |
## Bird), data = learning, family = poisson(link = log), na.action = "na.fail")
## ---
## Model selection table
## (Intrc) Group Sex TrlNm df logLik AICc delta weight
## 5 1.481 -0.05095 3 -2297.158 4600.4 0.00 0.511
## 7 1.509 + -0.05095 4 -2297.057 4602.2 1.82 0.205
## 6 1.456 + -0.05095 4 -2297.071 4602.2 1.85 0.202
## 8 1.485 + + -0.05095 5 -2296.960 4604.0 3.66 0.082
## 1 1.258 2 -2314.405 4632.8 32.47 0.000
## 3 1.286 + 3 -2314.303 4634.6 34.29 0.000
## 2 1.234 + 3 -2314.318 4634.7 34.32 0.000
## 4 1.262 + + 4 -2314.207 4636.5 36.12 0.000
## Models ranked by AICc(x)
## Random terms (all models):
## '1 | Bird'est.output<-model.avg(output, subset= delta < 2, revised.var = TRUE)
summary(est.output)##
## Call:
## model.avg(object = output, subset = delta < 2, revised.var = TRUE)
##
## Component model call:
## glmer(formula = NumErrors ~ <3 unique rhs>, data = learning, family =
## poisson(link = log), na.action = na.fail)
##
## Component models:
## df logLik AICc delta weight
## 3 3 -2297.16 4600.36 0.00 0.56
## 23 4 -2297.06 4602.18 1.82 0.22
## 13 4 -2297.07 4602.21 1.85 0.22
##
## Term codes:
## Group Sex TrialNumber
## 1 2 3
##
## Model-averaged coefficients:
## (full average)
## Estimate Std. Error Adjusted SE z value Pr(>|z|)
## (Intercept) 1.481487 0.086937 0.087100 17.009 <0.0000000000000002
## TrialNumber -0.050947 0.008641 0.008658 5.884 <0.0000000000000002
## SexM -0.013220 0.066629 0.066741 0.198 0.843
## GroupExperimental 0.012133 0.065997 0.066111 0.184 0.854
##
## (Intercept) ***
## TrialNumber ***
## SexM
## GroupExperimental
##
## (conditional average)
## Estimate Std. Error Adjusted SE z value Pr(>|z|)
## (Intercept) 1.481487 0.086937 0.087100 17.009 <0.0000000000000002
## TrialNumber -0.050947 0.008641 0.008658 5.884 <0.0000000000000002
## SexM -0.059137 0.130932 0.131188 0.451 0.652
## GroupExperimental 0.055070 0.131930 0.132188 0.417 0.677
##
## (Intercept) ***
## TrialNumber ***
## SexM
## GroupExperimental
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1ggplot(data=learning)+
geom_beeswarm(aes(x=TrialNumber,y=NumErrors),size=0.005, cex=0.1)+
geom_smooth(aes(x=TrialNumber, y=NumErrors), col="grey4", method="lm")+
scale_x_continuous(expand=c(0,0),name="Trial Number", limits=c(0,9), breaks=seq(1,8,1))+
scale_y_continuous(name="Number of Errors", limits=c(-2,max(learning$NumErrors)),breaks=seq(0,45,5), expand=c(0,0))+
theme(text=element_text(family="Calibri"))+
guides(col=F)+
theme_classic()
## Did birds vary in orientation strategy? ##
orient<- read.csv("mazeRotationResults.csv")
g1 <- glm(binStrat~Sex*Group, family=binomial(link="logit"),data=orient, na.action="na.fail")
summary(g1)##
## Call:
## glm(formula = binStrat ~ Sex * Group, family = binomial(link = "logit"),
## data = orient, na.action = "na.fail")
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -1.2278 -0.6039 -0.5701 1.1278 1.9479
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -1.6094 0.5477 -2.938 0.0033 **
## SexM -0.1252 0.8319 -0.150 0.8804
## GroupExperimental 1.7272 0.7322 2.359 0.0183 *
## SexM:GroupExperimental 0.1252 1.0790 0.116 0.9077
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 97.852 on 77 degrees of freedom
## Residual deviance: 85.552 on 74 degrees of freedom
## AIC: 93.552
##
## Number of Fisher Scoring iterations: 3dredge(g1)## Global model call: glm(formula = binStrat ~ Sex * Group, family = binomial(link = "logit"),
## data = orient, na.action = "na.fail")
## ---
## Model selection table
## (Int) Grp Sex Grp:Sex df logLik AICc delta weight
## 2 -1.6650 + 2 -42.787 89.7 0.00 0.684
## 4 -1.6420 + + 3 -42.783 91.9 2.16 0.233
## 8 -1.6090 + + + 4 -42.776 94.1 4.37 0.077
## 1 -0.7514 1 -48.926 99.9 10.17 0.004
## 3 -0.7673 + 2 -48.924 102.0 12.27 0.001
## Models ranked by AICc(x)#summary top model - treatment only
summary(glm(binStrat~Group, family=binomial(link="logit"),data=orient))##
## Call:
## glm(formula = binStrat ~ Group, family = binomial(link = "logit"),
## data = orient)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -1.2278 -0.5887 -0.5887 1.1278 1.9174
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -1.6650 0.4122 -4.040 0.0000535 ***
## GroupExperimental 1.7828 0.5366 3.322 0.000892 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 97.852 on 77 degrees of freedom
## Residual deviance: 85.574 on 76 degrees of freedom
## AIC: 89.574
##
## Number of Fisher Scoring iterations: 3## Plots ##
ggplot(data=orient, aes(x=Group, fill=Strat))+
geom_bar(position="dodge", size=0)+
scale_fill_manual(values=c("royalblue","gold3"))+
scale_y_continuous(expand=c(0,0), limits=c(0,50))+
#facet_grid(cols=vars(Group))+
xlab("Condition")+
ylab("Number of Birds") +
theme_classic() +
theme(legend.position="none",text=element_text(family="Calibri"))
orient2 <- orient[orient$Group=="Experimental",]
ggplot(data=orient2, aes(x=Diff, fill=Strat))+
geom_histogram(breaks=seq(-10,30,by=1), size=0)+
scale_fill_manual(values=c("royalblue","gold3"))+
scale_y_continuous(expand=c(0,0), limits=c(0,8), breaks=seq(0,8,1))+
scale_x_continuous(expand=c(0,0), limits=c(-10,35), breaks=seq(-10,30,5))+
xlab("Difference in Errors \n(Probe Trial - Final Training Trial)")+
ylab("Number of Birds") +
theme_classic() +
theme(legend.position="none",text=element_text(family="Calibri"))
Habitat Selection analysis
# Habitat selection
set.seed(106)
all_RSS <- read.csv("habitatOrientation_coefs.csv") #coefs and logrss values for each ind
all_RSS$strategy <- factor(all_RSS$strategy, levels= c("Egocentric", "Allocentric")) # change reference level to get allocentric estimates instead of mixed/ego birds. Clearer to explain.
all_avail <- read.csv("habitatOrientation_avail.csv") #availability of habitats for each ind
# selection (RSS)
rss_data <- all_RSS[all_RSS$key=="log_RSS_wood2other",]
rss_data <- merge(rss_data, all_avail, by = "id")
rss_data$fig_strategy <- factor(rss_data$strategy, levels=c("Allocentric", "Egocentric"))
ggplot(data= rss_data, aes(x=fig_strategy, y=mean, col=strategy)) +
geom_boxplot(show.legend = F, outlier.shape = NA)+
geom_pointrange(aes(ymin = lq, ymax = uq, alpha = inv_se), size=0.3,
position=position_jitterdodge(), show.legend = F) +
scale_color_manual(values=c( "gold3","royalblue"))+
theme_classic()+
theme(text=element_text(family="Calibri"), axis.text.x = element_text(size=7))+
scale_x_discrete(labels=c("Allocentric" = "Allocentric", "Egocentric" = "Mixed/Egocentric"))+
geom_hline(yintercept = 0, lty = 2) +
labs(x = "Strategy", y = "log-RSS")
habitat_glm <- glm(mean ~ sex * strategy + Avail_Habitat_other, data = rss_data, weight = inv_se, na.action="na.fail")
habitat_output <- dredge(habitat_glm, fixed = c("Avail_Habitat_other"))
habitat_output## Global model call: glm(formula = mean ~ sex * strategy + Avail_Habitat_other, data = rss_data,
## weights = inv_se, na.action = "na.fail")
## ---
## Model selection table
## (Int) Avl_Hbt_oth sex str sex:str df logLik AICc delta weight
## 1 -0.5593 0.00001277 3 14.510 -21.5 0.00 0.564
## 3 -0.5830 0.00001251 + 4 15.243 -19.8 1.70 0.241
## 2 -0.5647 0.00001227 + 4 14.621 -18.6 2.95 0.129
## 4 -0.5984 0.00001150 + + 5 15.636 -17.0 4.53 0.058
## 8 -0.5809 0.00001163 + + + 6 15.747 -13.0 8.49 0.008
## Models ranked by AICc(x)habitat_est.output <- model.avg(habitat_output, subset= delta < 2, revised.var = TRUE)
summary(habitat_est.output)##
## Call:
## model.avg(object = habitat_output, subset = delta < 2, revised.var = TRUE)
##
## Component model call:
## glm(formula = mean ~ <2 unique rhs>, data = rss_data, weights = inv_se,
## na.action = na.fail)
##
## Component models:
## df logLik AICc delta weight
## 1 3 14.51 -21.52 0.0 0.7
## 12 4 15.24 -19.82 1.7 0.3
##
## Term codes:
## Avail_Habitat_other strategy
## 1 2
##
## Model-averaged coefficients:
## (full average)
## Estimate Std. Error Adjusted SE z value
## (Intercept) -0.566382312 0.052888617 0.056612392 10.005
## Avail_Habitat_other 0.000012691 0.000002836 0.000003044 4.170
## strategyAllocentric 0.015152736 0.033633985 0.035007083 0.433
## Pr(>|z|)
## (Intercept) < 0.0000000000000002 ***
## Avail_Habitat_other 0.0000305 ***
## strategyAllocentric 0.665
##
## (conditional average)
## Estimate Std. Error Adjusted SE z value
## (Intercept) -0.566382312 0.052888617 0.056612392 10.005
## Avail_Habitat_other 0.000012691 0.000002836 0.000003044 4.170
## strategyAllocentric 0.050629435 0.044538001 0.047943212 1.056
## Pr(>|z|)
## (Intercept) < 0.0000000000000002 ***
## Avail_Habitat_other 0.0000305 ***
## strategyAllocentric 0.291
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1What about that outlier that does not select for woodland?
rss_data2 <- rss_data[rss_data$mean<0,]
ggplot(data= rss_data2, aes(x=fig_strategy, y=mean, col=strategy)) +
geom_boxplot(show.legend = F, outlier.shape = NA)+
geom_pointrange(aes(ymin = lq, ymax = uq, alpha = inv_se), size=0.3,
position=position_jitterdodge(), show.legend = F) +
scale_color_manual(values=c( "gold3","royalblue"))+
theme_classic()+
theme(text=element_text(family="Calibri"), axis.text.x = element_text(size=7))+
scale_x_discrete(labels=c("Allocentric" = "Allocentric", "Egocentric" = "Mixed/Egocentric"))+
geom_hline(yintercept = 0, lty = 2) +
labs(x = "Strategy", y = "log-RSS", tag="a")
habitat_glm <- glm(mean ~ sex * strategy + Avail_Habitat_other, data = rss_data2, weight = inv_se, na.action="na.fail")
habitat_output <- dredge(habitat_glm, fixed = c("Avail_Habitat_other"))
habitat_output## Global model call: glm(formula = mean ~ sex * strategy + Avail_Habitat_other, data = rss_data2,
## weights = inv_se, na.action = "na.fail")
## ---
## Model selection table
## (Int) Avl_Hbt_oth sex str sex:str df logLik AICc delta weight
## 1 -0.5780 0.00001370 3 20.380 -33.2 0.00 0.552
## 3 -0.5977 0.00001346 + 4 21.316 -31.8 1.39 0.276
## 2 -0.5800 0.00001350 + 4 20.409 -30.0 3.20 0.111
## 4 -0.6065 0.00001285 + + 5 21.563 -28.5 4.65 0.054
## 8 -0.5926 0.00001295 + + + 6 21.688 -24.4 8.78 0.007
## Models ranked by AICc(x)habitat_est.output <- model.avg(habitat_output, subset= delta < 2, revised.var = TRUE)
summary(habitat_est.output)##
## Call:
## model.avg(object = habitat_output, subset = delta < 2, revised.var = TRUE)
##
## Component model call:
## glm(formula = mean ~ <2 unique rhs>, data = rss_data2, weights =
## inv_se, na.action = na.fail)
##
## Component models:
## df logLik AICc delta weight
## 1 3 20.38 -33.16 0.00 0.67
## 12 4 21.32 -31.77 1.39 0.33
##
## Term codes:
## Avail_Habitat_other strategy
## 1 2
##
## Model-averaged coefficients:
## (full average)
## Estimate Std. Error Adjusted SE z value
## (Intercept) -0.584529560 0.040240575 0.043229689 13.521
## Avail_Habitat_other 0.000013623 0.000002142 0.000002308 5.901
## strategyAllocentric 0.014253783 0.027821499 0.028922734 0.493
## Pr(>|z|)
## (Intercept) <0.0000000000000002 ***
## Avail_Habitat_other <0.0000000000000002 ***
## strategyAllocentric 0.622
##
## (conditional average)
## Estimate Std. Error Adjusted SE z value
## (Intercept) -0.584529560 0.040240575 0.043229689 13.521
## Avail_Habitat_other 0.000013623 0.000002142 0.000002308 5.901
## strategyAllocentric 0.042749291 0.033215808 0.035926358 1.190
## Pr(>|z|)
## (Intercept) <0.0000000000000002 ***
## Avail_Habitat_other <0.0000000000000002 ***
## strategyAllocentric 0.234
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1# speed
speed_data <- all_RSS[all_RSS$key%in% c("mean_speed_wood", "mean_speed_other"),]
speed_data$key <- factor(speed_data$key, levels=c("mean_speed_wood", "mean_speed_other"), labels = c("Wood", "Other"))
ggplot(data= speed_data, aes(x=sex, y=mean, col=key)) +
geom_boxplot()+
geom_pointrange(aes(ymin = lq, ymax = uq, alpha = inv_se),
size = 0.3, position=position_jitterdodge(), show.legend = F) +
scale_color_manual(values=c("forestgreen", "purple"))+
theme_classic()+
theme(legend.position = "none",text=element_text(family="Calibri"))+
scale_y_continuous(limits=c(0,60), expand= c(0,0))+
guides(col = guide_legend(title="Habitat"))+
labs(x = "Sex", y = "Mean Displacement Distance (m/5min)")
speed_glm <- glm(mean~ sex * strategy * key, data=speed_data, weight= inv_se,na.action="na.fail")
speed_output <- dredge(speed_glm)
speed_output #only 1 model## Global model call: glm(formula = mean ~ sex * strategy * key, data = speed_data,
## weights = inv_se, na.action = "na.fail")
## ---
## Model selection table
## (Int) key sex str key:sex key:str sex:str key:sex:str df logLik AICc
## 12 20.91 + + + 5 -67.565 146.9
## 16 21.00 + + + + 6 -67.447 149.4
## 48 21.12 + + + + + 7 -67.296 152.1
## 32 21.02 + + + + + 7 -67.431 152.4
## 4 20.60 + + 4 -72.310 153.8
## 64 21.12 + + + + + + 8 -67.293 155.2
## 8 20.70 + + + 5 -72.195 156.2
## 128 20.99 + + + + + + + 9 -66.593 157.2
## 40 20.86 + + + + 6 -71.987 158.5
## 24 20.68 + + + + 6 -72.173 158.9
## 56 20.83 + + + + + 7 -71.928 161.4
## 2 21.54 + 3 -82.644 172.0
## 6 21.73 + + 4 -82.291 173.7
## 22 21.70 + + + 5 -82.272 176.3
## 3 21.76 + 3 -110.359 227.4
## 7 21.72 + + 4 -110.355 229.9
## 39 21.84 + + + 5 -110.336 232.4
## 1 23.36 2 -114.727 233.8
## 5 23.48 + 3 -114.701 236.1
## delta weight
## 12 0.00 0.672
## 16 2.55 0.188
## 48 5.20 0.050
## 32 5.47 0.044
## 4 6.87 0.022
## 64 8.34 0.010
## 8 9.26 0.007
## 128 10.29 0.004
## 40 11.63 0.002
## 24 12.00 0.002
## 56 14.46 0.000
## 2 25.06 0.000
## 6 26.83 0.000
## 22 29.42 0.000
## 3 80.49 0.000
## 7 82.96 0.000
## 39 85.54 0.000
## 1 86.88 0.000
## 5 89.17 0.000
## Models ranked by AICc(x)speed_glm <- glm(mean~ sex * key, data=speed_data, weight= inv_se,na.action="na.fail")
summary(speed_glm)##
## Call:
## glm(formula = mean ~ sex * key, data = speed_data, weights = inv_se,
## na.action = "na.fail")
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -119.334 -33.184 3.744 29.514 146.101
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 20.9135 0.2730 76.614 < 0.0000000000000002 ***
## sexM 1.2174 0.3814 3.192 0.00293 **
## keyOther 4.1919 0.6057 6.921 0.0000000417 ***
## sexM:keyOther 2.3189 0.7469 3.105 0.00370 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for gaussian family taken to be 3544.053)
##
## Null deviance: 1348702 on 39 degrees of freedom
## Residual deviance: 127586 on 36 degrees of freedom
## AIC: 145.13
##
## Number of Fisher Scoring iterations: 2#directionality
dir_data <- all_RSS[all_RSS$key%in% c("cos_ta_", "habitat_start_cos_ta"),]
dir_data$key <- factor(dir_data$key, levels=c("cos_ta_", "habitat_start_cos_ta"), labels = c("Wood", "Other"))
ggplot(data= dir_data, aes(x=sex, y=mean, col=key)) +
geom_boxplot()+
geom_pointrange(aes(ymin = lq, ymax = uq, alpha = inv_se),
size = 0.3, position=position_jitterdodge(), show.legend = F) +
scale_color_manual(values=c("forestgreen", "purple"))+
theme_classic()+
theme(legend.position = "none",text=element_text(family="Calibri"))+
guides(col = guide_legend(title="Habitat"))+
geom_hline(yintercept = 0, lty = 2) +
labs(x = "Sex", y = expression(paste("iSSA cos(turning angle) ", beta ," coefficient")))
dir_glm <- glm(mean~ strategy * key * sex, data=dir_data, weights = inv_se,na.action="na.fail")
dir_output <- dredge(dir_glm)
dir_output## Global model call: glm(formula = mean ~ strategy * key * sex, data = dir_data, weights = inv_se,
## na.action = "na.fail")
## ---
## Model selection table
## (Int) key sex str key:sex key:str sex:str key:sex:str df logLik AICc
## 4 -0.8168 + + 4 14.681 -20.2
## 12 -0.7974 + + + 5 15.311 -18.9
## 8 -0.8029 + + + 5 14.819 -17.9
## 16 -0.7834 + + + + 6 15.455 -16.4
## 40 -0.7733 + + + + 6 15.126 -15.7
## 24 -0.7965 + + + + 6 14.865 -15.2
## 48 -0.7540 + + + + + 7 15.768 -14.0
## 32 -0.7725 + + + + + 7 15.566 -13.6
## 3 -0.7785 + 3 9.741 -12.8
## 56 -0.7717 + + + + + 7 15.134 -12.8
## 64 -0.7496 + + + + + + 8 15.812 -11.0
## 7 -0.7650 + + 4 9.841 -10.5
## 39 -0.7351 + + + 5 10.085 -8.4
## 128 -0.7693 + + + + + + + 9 16.166 -8.3
## 2 -0.7064 + 3 6.117 -5.6
## 6 -0.6792 + + 4 6.585 -4.0
## 22 -0.6706 + + + 5 6.647 -1.5
## 1 -0.6260 2 0.257 3.8
## 5 -0.5963 + 3 0.691 5.3
## delta weight
## 4 0.00 0.434
## 12 1.36 0.220
## 8 2.35 0.134
## 16 3.86 0.063
## 40 4.51 0.045
## 24 5.03 0.035
## 48 6.18 0.020
## 32 6.59 0.016
## 3 7.40 0.011
## 56 7.45 0.010
## 64 9.24 0.004
## 7 9.68 0.003
## 39 11.81 0.001
## 128 11.89 0.001
## 2 14.65 0.000
## 6 16.19 0.000
## 22 18.69 0.000
## 1 24.03 0.000
## 5 25.50 0.000
## Models ranked by AICc(x)dir_est.output<-model.avg(dir_output, subset= delta < 2, revised.var = TRUE)
summary(dir_est.output)##
## Call:
## model.avg(object = dir_output, subset = delta < 2, revised.var = TRUE)
##
## Component model call:
## glm(formula = mean ~ <2 unique rhs>, data = dir_data, weights = inv_se,
## na.action = na.fail)
##
## Component models:
## df logLik AICc delta weight
## 12 4 14.68 -20.22 0.00 0.66
## 123 5 15.31 -18.86 1.36 0.34
##
## Term codes:
## key sex key:sex
## 1 2 3
##
## Model-averaged coefficients:
## (full average)
## Estimate Std. Error Adjusted SE z value Pr(>|z|)
## (Intercept) -0.81032 0.03977 0.04106 19.736 < 0.0000000000000002 ***
## keyOther 0.11620 0.06690 0.06863 1.693 0.090419 .
## sexM 0.18887 0.05173 0.05332 3.542 0.000397 ***
## keyOther:sexM 0.03409 0.07281 0.07426 0.459 0.646168
##
## (conditional average)
## Estimate Std. Error Adjusted SE z value Pr(>|z|)
## (Intercept) -0.81032 0.03977 0.04106 19.736 < 0.0000000000000002 ***
## keyOther 0.11620 0.06690 0.06863 1.693 0.090419 .
## sexM 0.18887 0.05173 0.05332 3.542 0.000397 ***
## keyOther:sexM 0.10147 0.09457 0.09786 1.037 0.299758
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1