First at all, thank you for your excellent work.

I made the next code sjt.frq(data$x, sort.frq = "asc") and one of the levels (Mañana) is transform with NA.

I thank you if this issue can be fix.

Thank in advanced and congratulations for the great package

Example:

testdata <- structure(c(7L, 8L, 8L, 7L, 7L, 7L, 8L, 7L, 7L, 7L, 7L, 8L, 7L, 
                        7L, 7L, 7L, 7L, 9L, 7L, 8L, 8L, 8L, 7L, 7L, 7L, 7L, 7L, 8L, NA, 
                        8L, 8L, 4L, 8L, 7L, 8L, 7L, 8L, 7L, 8L, 7L, 7L, 7L, 7L, 8L, 6L, 
                        8L, 7L, 7L, 8L, 7L, 7L, 7L, 8L, 8L, 7L, NA, 7L, 7L, 7L, 7L, 7L, 
                        7L, 7L, 7L, 8L, 7L, 7L, 7L, 8L, 7L, 8L, 7L, 6L, 7L, 7L, 7L, 8L, 
                        8L, 7L, 8L, 8L, 4L, 8L, 7L, 8L, 8L, 8L, 8L, 7L, 7L, 7L, 8L, 8L, 
                        8L, 9L, 8L, 8L, 8L, 7L, 7L, 4L, 8L, 8L, 4L, 5L, 7L, 8L, 4L, 6L, 
                        7L, 6L, 4L, 4L, 6L, 4L, 7L, 7L, 7L, 8L, 8L, 4L, 4L, 4L, 4L, 4L, 
                        4L, 8L, 4L, 4L, 7L, 7L, 7L, 4L, 7L, 7L, NA, 4L, 8L, 4L, 4L, 8L, 
                        4L, 8L, 4L, 8L, 7L, 7L, 8L, 6L, 8L, 4L, 8L, 8L, 7L, 8L, 7L, 8L, 
                        8L, 8L, 8L, 4L, 8L, 6L, 6L, 4L, 7L, 8L, 7L, 8L, 7L, 7L, 4L, 8L, 
                        3L, 6L, 6L, 8L, 7L, 7L, 4L, 3L, 6L, 4L, 6L, 4L, 7L, NA, 8L, 4L, 
                        4L, 8L, 6L, 8L, 8L, 7L, 8L), .Label = c("Noch Schüler", "Schule beendet ohne Abschluss", 
                                                                "Hauptschulabschluss/Volksschulabschluss", "Realschulabschluss (Mittlere Reife)", 
                                                                "Abschluss Polytechnische Oberschule 10. Klasse (vor 1965: 8. Klasse)", 
                                                                "Fachhochschulreife (Abschluss einer Fachoberschule)", "Abitur, allgemeine oder fachgebundene Hochschulreife (Gymnasium bzw. EOS)", 
                                                                "Hochschulabschluss", "Anderer Schulabschluss:"), class = "factor")

sjt.frq(testdata)

prints:

yet

table(testdata)

shows the above table is wrong:

                                                         Noch Schüler 
                                                                    0 
                                        Schule beendet ohne Abschluss 
                                                                    0 
                              Hauptschulabschluss/Volksschulabschluss 
                                                                    2 
                                  Realschulabschluss (Mittlere Reife) 
                                                                   31 
 Abschluss Polytechnische Oberschule 10. Klasse (vor 1965: 8. Klasse) 
                                                                    1 
                  Fachhochschulreife (Abschluss einer Fachoberschule) 
                                                                   13 
Abitur, allgemeine oder fachgebundene Hochschulreife (Gymnasium bzw. EOS) 
                                                                   77 
                                                   Hochschulabschluss 
                                                                   66 
                                              Anderer Schulabschluss: 
                                                                    2 

Any hints what causes this?

Siehe weightby in sjp.stackfrq

"emm" with lmer fails for me, if I use lmerTest I get

Error in summary(fit)$coefficients[-1, 4] : subscript out of bounds

and with plain lme4

Error in colnames<-(tmp, value = c("x", "y", "grp", "l.ci", "u.ci", :
'names' attribute [6] must be the same length as the vector [5]

Variables in my dataset do not have incremental levels (e.g. 1, 2, 5, 15, 20, 35, 99) therefore, I have problems with obtaining the right result in sjt.* functions.

Please find example code for clarification below

library(sjPlot)

### example data
example_data <- data.frame(labelled(
  c(10,20,30,10,20,30,10,99,30,20),
  c('c'=10,'b'=20,'a'=30,'NA'=99)))

names(example_data) <- 'V1'

## incorrect result (only first group is visible)
sjt.frq(example_data$V1) 

## correct results (however without labels)
sjt.frq(as.numeric(example_data$V1)) 

sample file:
https://drive.google.com/open?id=0B8FLxb6Zof5udUNnSTY4NXZvN2c&authuser=0

raw.data3 <- read_spss(
  "multibyte_char_data.sav",
  enc = "UTF-8", 
  autoAttachVarLabels = T, 
  option = "foreign")

view_spss(raw.data3, encoding = "UTF-8")

Session info ---------------------------------------------------
  setting  value                       
  version  R version 3.1.3 (2015-03-09)
  system   i386, mingw32               
  ui       RStudio (0.99.380)          
  language (EN)                        
  collate  Chinese_Taiwan.950          
  tz       Asia/Taipei  

Hi, I am trying to make an interaction plot for a logit model (binary dependent variable) where both independent and moderating variable are continuous. I use meansd to get three curves for moderation. However, I keep getting the following error message:

Error in contrasts<-(*tmp*, value = contr.funs[1 + isOF[nn]]) :
contrasts can be applied only to factors with 2 or more levels
In addition: Warning messages:
1: In mean.default(X[, name]) :
argument is not numeric or logical: returning NA
2: In mean.default(X[, name]) :
argument is not numeric or logical: returning NA

Tried to google, but still not found a proper solution. (Still rooky with R).

Any inputs? Thanks.

Label Color wird bei theme presets überschrieben, NULL Check vorher

I have a model with the following formula:

model_LH<- glmer(Tipo ~ Co.2.Pr + Education + Negation + Pr.2.Pr + Typical.Action.Chain.Pos + Style*Tense +(1 + Co.2.Pr | Muestra)+ (1 + Negation | Muestra) +(1 + Typical.Action.Chain.Pos| Muestra)+ (1 | Nombrelema), data=datalh, family="binomial", control=glmerControl(optimizer="bobyqa"))

When I run
sjp.glmer(model_LH, type = "ri.pc", show.se = TRUE)

I get the following error:
Error in $<-.data.frame(*tmp*, "xbeta", value = numeric(0)) :
replacement has 0 rows, data has 2093

Really love the sjt.int() function. One suggestion I have is for sjt.int() to recognise I() in a formula.

[example is from Hayes (2013) great book about moderation (and mediation)]

h31<- lm(liking ~ protest + sexism + protest*sexism,    data=protest) 
h32<- lm(liking ~ protest + sexism + I(protest*sexism), data=protest)

sjp.int(h31) # works
sjp.int(h32) # does not work

I am using the latest version from github as of today

Is there a function to plot the logistf?

I have lme4 installed and loaded, but sjp.lmer() gives this message:

Error: Package 'lme4' needed for this function to work. Please install it.

my model with lmer function

model_2 = lmer (systolic_BP ~ time2 + (1 + time2 | ID), REML = FALSE, data = subset_data)
summary (model_2)

I recieve the following error although "lme4" is installed and running

sjp.lmer(model_2)
Error: Package 'lme4' needed for this function to work. Please install it.

Hi,

When I use the write_spss function with variable.label attributes, the labels appear to export properly into the .sav, however they are cut off after 120 characters. I believe 120 characters was the limit in previous versions of SPSS, but I was wondering if there was a way to change this limit. The labels I need are somewhat lengthy survey questions, so it's important the text isn't cut off after 120 characters, and I've had trouble finding help for this issue elsewhere.

Thanks for the great package.

Seems like it should not be a problem, but it is.

Hi,
Would it be possible to pull-out the actual p-value from the lm/glm result? So, when p <0.001, not <0.001, but for instance p = 0.00045637, or p = 4.56e-5?

Would be great. Or is this already possible?

Thanks!

Sander

APA styled p-values not in sjt.xtab.

R version 3.2, latest sjPlot (1.8) from the repository. Following your example use of sjp.glmer(), it reports that lme4 is not installed, although lme4 is both installed and functioning:

# fit model
library(lme4)
Loading required package: Matrix
Loading required package: Rcpp
# for "dicho" function etc.
library(sjmisc)
# create binary response
sleepstudy$Reaction.dicho <- dicho(sleepstudy$Reaction, dichBy = "md")
# fit model
fit <- glmer(Reaction.dicho ~ Days + (Days | Subject),
             sleepstudy,
             family = binomial("logit"))
sjp.glmer(fit)
Error: Package 'lme4' needed for this function to work. Please install it.

I installed the latest development snapshot.
library(devtools)
devtools::install_github("sjPlot/devel")

And then trying to load a dataset by read_spss. I got this error:
Error: could not find function "read_spss"

But ?read_spss works well.

If I reinstall the package using install.packages("sjPlot"). Everything looks fine.

If the count of the first bar of a frequency plot is 0, wrong values are assigned to the plots.
Two reproducible examples with different results, depending on how the level order is assigned.

mydf <- data.frame(Q=c(rep("No", 5), rep("Maybe", 3)))
levels(mydf[,1]) <- c("Yes", "No", "Maybe")
sjp.frq(mydf[,1], axisLimits.y=c(0,7))

mydf <- data.frame(Q=c(rep("No", 5), rep("Maybe", 3)))
mydf[,1] <- ordered(mydf[,1], levels=c("Yes", "No", "Maybe"))
sjp.frq(mydf[,1], axisLimits.y=c(0,7))

sjp.likert(df) returns an error if ncol(df)=1.

library(sjPlot)
set.seed(77)
df <- data.frame(Q1=factor(sample(1:4, 100, replace=TRUE)), Q2=factor(sample(1:4, 100, replace=TRUE)))
sjp.likert(df) # works fine!
sjp.likert(df[, 1, drop=FALSE]) # gives an error!

Would be nice if sjt.corr could display different correlation coefficients in the upper and lower triangle of the matrix. I think it is quite common to display Pearson and Spearman correlation coefficients and to do this in one matrix.

lower.tri(x, diag = FALSE) and upper.tri(x, diag = FALSE) might come in handy as mentioned here:
http://www.sthda.com/english/wiki/elegant-correlation-table-using-xtable-r-package

if showType=T, date variables are presented as atomic

Predictor labels also require name for intercept, which is confusing.

Would it be possible to have axisLimits.x parameter in the sjp.int plot (and other plots as well)?

Hello Daniel,

I'm using sjp.likert function from de 'sjPlot' package, to make a series of plots of different dataframes in which their columns correspond to different questions that follow a likert format.

The exact dataframe is reproduced below:

col_1 <- c(2,1,1,5)
col_2 <- c(2,1,1,2)
col_3 <- c(2,1,1,2)
col_4 <- c(2,1,1,2)
col_5 <- c(2,1,1,5)

df <- as.data.frame(cbind(col_1,col_2,col_3,col_4,col_5))

Following the instructions of the 'sjPlot' package, I gave labels to the questions as well as to the possible answers(value labels).

question.labels <- c("Las personas de mi Equipo están calificadas adecuadamente para desempeñar su trabajo", "Mi equipo es eficiente para solucionar problemas sin perder tiempo en encontrar culpables", "Mi área busca formas de hacer los procesos de manera inteligente y eficiente", "El ambiente en mi área es generalmente bueno", "En mi Equipo tenemos una dinámica de trabajo que permite nuestro mejor desempeño")

value.labels <- c("strongly agree", "agree", "disagree", "strongly disagree", "neither agree or disagree")

Then I use the sjp.likert function to produce the plot:

sjp.likert(bbdd_mod[, lik.f2],cat.neutral = 5, axisLabels.y = items[lik.f2-7], legendLabels = labels, value.labels = "sum.outside", gridRange = 1.4)

But I get the following error:

Error in data.frame(..., check.names = FALSE) : arguments imply differing number of rows: 5, 3

While with the following dataframe there is no problem:

col_a <- c(2,3,1,2,2,1)
col_b <- c(2,4,2,5,2,2)
col_c <- c(2,3,2,2,2,2)
col_d <- c(2,2,1,2,2,1)
col_e <- c(3,5,1,2,2,1)

df2 <- cbind(col_a,col_b,col_c,col_d,col_e)

In df2 all the possible values are present, from 1 to 5, that is the only difference with df, in which only 1,2 and 5 are possible values. Hence I thought this could explain the error, that there are 3 possible values in the data frame while the value.labels contain 5 distinct options. But even when I take out the argument legendLabels from the function (so there is no incongruency between the actual values and all the possible labels) I still get the same error when trying to plot df.

Is there a way to solve this?

Thanks in advance

Quick example

library(survey)
data(api)
dstrat<-svydesign(id=~1,strata=~stype, weights=~pw, data=apistrat, fpc=~fpc)

fit1<-svyglm(sch.wide~ell+meals+mobility, design=dstrat,family = binomial(link = "logit"))
fit2<-svyglm(sch.wide~ell+meals*mobility, design=dstrat,family = binomial(link = "logit"))

sjt.glm(fit1, fit2)

Please consider the following:

> P.min <- P[1:10,1:10]
> P.min
   6036 6053 6054 6037 6055 6056 6057 6058 6059 6060
1     1    1    1    1    1    1    1    1    1    1
2     0    1    1    1    1    1    1    1    0    1
3     1    0    1    1    1    1    1    1    1    1
4     0    1    1    0    1    0    0    1    1    1
5     1    1    1    1    1    1    1    1    1    1
6     0    1    1    1    1    1    1    1    1    1
7     0    0    1    0    0    1    1    1    1    1
8     1    1    1    0    1    1    1    1    1    1
9     1    1    1    1    0    0    1    1    1    1
10    1    0    0    1    0    1    0    0    1    0

The result is

  |======================================================================| 100%
Error in gsub("      <", "<", html.table, fixed = TRUE) : 
  object 'toWrite' not found

I have attempted the same with larger tables. P.min is an arbitrary subset of a larger table (36 columns, 2466 rows).

Any clues?

Much appreciated,

• roffe

To plot each single predictor of fitted model against predicted values of the outcome. May be used to check for polynomial terms, whether they fit well or not. Loess-smoothing possible, but probably no CI.

no text

It might be useful to add an option to plot the sum of fixed effect and random effect. In other words, it would allow the random effect to be plotted around the fixed effect instead of just zero.

This works:

fit <- lmer(Reaction ~ Days + (Days | Subject), sleepstudy)
sjp.lmer(fit)

This doesNOT work

fit <- lmer(Reaction ~ Days + (Days | | Subject), sleepstudy)
sjp.lmer(fit)

Could you check if sjp.int( , moderatorValue="minmax") calculates correctly?
Just checked the plots with the dataset in Hayes (2013), p. 232 [dataset protest is available at http://afhayes.com/introduction-to-mediation-moderation-and-conditional-process-analysis.html], however, I think you can follow without this specific dataset.
I have:

head(protest)
Source: local data frame [6 x 4]

sexism liking respappr protest
1 4.87 4.83 4.25 1
2 4.25 4.50 5.75 0
3 5.00 5.50 4.75 1
4 5.50 5.66 7.00 1
5 5.62 6.16 6.75 1
6 5.75 6.00 5.50 1

h3 <- lm(liking ~ protest + sexism + protest:sexism, data=protest)
h3
Call:
lm(formula = liking ~ protest + sexism + protest:sexism, data = protest)

Coefficients:
(Intercept) protest sexism protest:sexism
7.7062 -3.7727 -0.4725 0.8336

this should replicate the minmax calculation of sjp.int()

pro2 <- c(0,1,0,1)
sex2 <- c(rep(min(protest$sex), 2), rep(max(protest$sexism), 2))
df2 <- data.frame(protest=pro2, sexism=sex2)
df2 <- cbind(df2, pred = predict(h3, df2))

df2
protest sexism pred
1 0 2.87 6.350185
2 1 2.87 4.969828
3 0 7.00 4.398798
4 1 7.00 6.461013

compare the predicted (fitted) values to sjp.int's plot; sjp.int's maximum values seem to be off. The lines should cross each other at a value of sexim (moderator) of slightly above 4.5

sjp.int(h3, moderatorValue="minmax", showValueLabels = TRUE)

When I used view_spss to create a variable list,
I saw that those items with 0 respondents won't shows "0" in freq and %; the empty cells were replaced by latter frequencies. In my case,

the 5th and 6th items have 0 responses, but the two cells that are supposed to 0 were replaced by later figures. Please advice. Thank you!

sessionInfo()
R version 3.1.3 (2015-03-09)
Platform: x86_64-apple-darwin13.4.0 (64-bit)
Running under: OS X 10.10.3 (Yosemite)

I get the following error message with type='emm':

library(lme4); library(lmerTest); library(effects); library(sjPlot)
lme4::lmer(`Short term attractiveness` ~ Cue * gFaktor_Mann  +  (1|ID_Frau), data = comb) -> m1
sjp.int(fit = m1, plevel = 1,type ="emm")

Computing approximate p-values via Wald chi-squared test...
Error in colnames<-(*tmp*, value = c("x", "y", "grp", "l.ci", "u.ci", :
'names' attribute [6] must be the same length as the vector [5]

Thank you very much for your work, this is useful.

I am trying to use sjp.glmer from the version I just installed from github. I tried your example from: http://www.r-bloggers.com/visualizing-generalized-linear-mixed-effects-models-with-ggplot-rstats-lme4/

library(sjPlot)
library(lme4)
fit <- lmer(Reaction ~ Days + (Days | Subject), sleepstudy)

simple plot

sjp.lmer(fit)

This gives me the following error:

sjp.lmer(fit)
Error in array(NA, c(J, K)) : 'dims' cannot be of length 0

Please could you advise if there is a way I could use the function?

Thank you very much
Dominik

I've fitted a series of models and am looking to use your very useful package to create some diagnostic plots (thanks for writing, maintaining and updating by the way).

One of my returned models is stored in t$itt and I can plot the fixed effects without any problem, but trying to plot the standardised version of the fixed effects results in the error 'object 'fit' not found'...

sjp.lmer(t$itt, type = 'fe')
sjp.lmer(t$itt, type = 'fe.std')
Error in lme4::fixef(fit) : object 'fit' not found

Also trying to plot type = 'fe.prob' I get the following error...

sjp.lmer(t$itt, type = "fe.prob")
Error in $<-.data.frame(*tmp*, "x", value = c(2L, 1L)) :
replacement has 2 rows, data has 0

A minimal reproducible example of both of these errors can be done using the examples you kindly provide on your web-page (http://www.strengejacke.de/sjPlot/sjp.lmer/)...

data(efc)
efc$grp = as.factor(efc$e15relat)
levels(x = efc$grp) <- sji.getValueLabels(efc$e15relat)
mydf <- na.omit(data.frame(neg_c_7 = as.numeric(efc$neg_c_7),
sex = as.factor(efc$c161sex),
c12hour = as.numeric(efc$c12hour),
barthel = as.numeric(efc$barthtot),
grp = efc$grp))
mydf <- na.omit(data.frame(neg_c_7 = as.numeric(efc$neg_c_7),
sex = as.factor(efc$c161sex),
c12hour = as.numeric(efc$c12hour),
barthel = as.numeric(efc$barthtot),
grp = efc$grp))
fit2 <- lmer(neg_c_7 ~ sex + c12hour + barthel + (1|grp),
data = mydf)
sjp.lmer(fit2, type = "fe")
sjp.lmer(fit2, type = "fe.std")
Error in lme4::fixef(fit) : object 'fit' not found
sjp.lmer(fit2, type = "fe.prob")
Error in $<-.data.frame(*tmp*, "x", value = c(2L, 1L)) :
replacement has 2 rows, data has 0

I'm using the most recent update of sjPlot_1.6.8 as my sessionInfo() shows...

sessionInfo()
R version 3.1.2 (2014-10-31)
Platform: x86_64-pc-linux-gnu (64-bit)
locale:
[1] C
attached base packages:
[1] datasets grid grDevices splines graphics utils stats
[8] methods base
other attached packages:
[1] texreg_1.34 sjPlot_1.6.8 ResourceSelection_0.2-4
[4] lme4_1.1-7 Rcpp_0.11.4 Matrix_1.1-5
[7] gridExtra_0.9.1 XML_3.98-1.1 shiny_0.11
[10] RSQLite_1.0.0 DBI_0.3.1 rms_4.2-1
[13] SparseM_1.6 reshape2_1.4.1 dplyr_0.4.1
[16] lubridate_1.3.3 knitr_1.9 Hmisc_3.14-6
[19] Formula_1.2-0 lattice_0.20-29 ggplot2_1.0.0
[22] epicalc_2.15.1.0 nnet_7.3-8 MASS_7.3-37
[25] survival_2.37-7 foreign_0.8-62
loaded via a namespace (and not attached):
[1] R6_2.0.1 RColorBrewer_1.1-2 RJSONIO_1.3-0
[4] TH.data_1.0-6 abind_1.4-0 acepack_1.3-3.3
[7] arm_1.7-07 assertthat_0.1 car_2.0-24
[10] cluster_2.0.1 coda_0.16-1 colorspace_1.2-4
[13] compiler_3.1.2 digest_0.6.8 evaluate_0.5.5
[16] formatR_1.0 gtable_0.1.2 htmltools_0.2.6
[19] httpuv_1.3.2 labeling_0.3 latticeExtra_0.6-26
[22] magrittr_1.5 memoise_0.2.1 mgcv_1.8-4
[25] mime_0.2 minqa_1.2.4 mnormt_1.5-1
[28] multcomp_1.3-8 munsell_0.4.2 mvtnorm_1.0-2
[31] nlme_3.1-119 nloptr_1.0.4 parallel_3.1.2
[34] pbkrtest_0.4-2 plyr_1.8.1 polspline_1.1.9
[37] proto_0.3-10 psych_1.5.1 quantreg_5.11
[40] rpart_4.1-8 sandwich_2.3-2 scales_0.2.4
[43] stringr_0.6.2 tools_3.1.2 xtable_1.7-4
[46] zoo_1.7-11

According to the documentation, the return value of sjp.glm is a 'structure' with the ggplot object in the plot 'element'.

This is not true when the type is set to 'vif'. But the resulting ggplot object should be returned here as well.

recoding a factor into 0/1 does it for me.

This was unclear to me from the docs and from the way the function behaves:
plevel does not make sense, if you permit only lme4::lmer (which does not give _p_s) and
showCI does not seem to do anything. I can share some code that I use for simple and bootstrapped CIs on predictions, if you're interested.

That concludes it, sorry for the many issues filed, really like the package, but these issues made for a discouraging experience for the student that I recommended the package to (not yet at the level to debug this sort of stuff).

The documentation of sjp.glm says

        • ‘"vif"’ to plot Variance Inflation Factors. See details.

But there is no 'details' section.

Just read that plm support has been added to sjt.int, great! It works for pooled OLS (pooling), but it does not seem to work for fixed effects ("within") nor random effects ("random").

pooled <- plm(y ~ a + b + a:b, data=pdat, model="pooling")
fe <- plm(y ~ a + b + a:b, data=pdat, model="within")
re <- plm(y ~ a + b + a:b, data=pdat, model="random")

sjp.int(pooled, plevel = 0.9) # works
sjp.int(fe, plevel = 0.9) # Error: Predictors of interaction terms (main effects) must be included as single term as well. See Note in ?sjp.int 

[same for re]

Let me know if you need further input.

Good morning, First all, thank for your great package.

Is it possible to extract the output of a sjPlot table function in a data.frame, for example:

library(sjmisc)
data(efc)
a<- sjt.frq(efc$e42dep)

Something like that

b<- sj-data.frame(a)

What can I do for solve the above issule?

Thank you in advance
José F. Zea

It seems that the effects package somehow hangs, when there are two random effects. sjPlot is indirectly affected by this, and I didn't quickly find out where to lodge bugs for effects.

Hi.

Now use this package with knitr is not comfortable. There is native knitr print method described in the vignette: vignette('knit_print', package = 'knitr').

It would be nice if you'll realize this.

You might want to have a look at the probemod package for printing (and calculation) of (among other) the Jason Neyman method for significant parts of the interaction. Plot is not as pretty as sjt.int's, though.

It would be really great if very small p-values that are being currently shown as "p=0.000" can be displayed everywhere as being "p<0.001" to follow the APA guidelines.

When feeding my logistic regression model into sjp.glm I see this:

> sjp.glm(model.logreg)
Waiting for profiling to be done...
Error in seq.default(lower_lim, upper_lim, by = gridBreaksAt) : 
  'by' argument is much too small

I tried to pass several values for gridBreaksAt from 0.01 to 100000000 to no avail.

The example from the help works fine.

Seems like sjt.corr does not display p values (nor indicates significance by asteriks) when a cor object is passed. It works with a data.frame passed. I am using version 1.8 from CRAN.

df <- as.data.frame(cbind(rnorm(10),
                          rnorm(10),
                          rnorm(10),
                          rnorm(10),
                          rnorm(10)))

# significance is indicated
sjt.corr(df) 

cor_object <- cor(df)

# significance neither indicated nor p values printed
sjt.corr(cor_object)
sjt.corr(cor_object,
         showPValues = TRUE)