psych::alpha()

2017-09-12 6 min read rstats

This is a tutorial in progress on how to calculate Cronbach’s alphas using the psych package in R.

library(tidyverse)
library(psych)

disgust <- read_csv("https://psyteachr.github.io/msc-data-skills/data/disgust.csv")

SPSS

  • Under the Analyze menu, choose Scale and Reliability Analysis...
  • Choose the 7 moral disgust items
  • Make sure the Model is “Alpha”
  • Under Statistics, add descriptives for Item, Scale, and Scale if Item Deleted

psych::alpha()

alpha(x, keys=NULL, cumulative=FALSE, title=NULL, max=10, na.rm=TRUE,
      check.keys=FALSE, n.iter=1, delete=TRUE, use="pairwise", warnings=TRUE, n.obs=NULL)

Arguments

  • x A data.frame or matrix of data, or a covariance or correlation matrix
  • keys If some items are to be reversed keyed, then either specify the direction of all items or just a vector of which items to reverse
  • title Any text string to identify this run
  • cumulative should means reflect the sum of items or the mean of the items. The default value is means.
  • max the number of categories/item to consider if reporting category frequencies. Defaults to 10, passed to link{response.frequencies}
  • na.rm The default is to remove missing values and find pairwise correlations
  • check.keys if TRUE, then find the first principal component and reverse key items with negative loadings. Give a warning if this happens.
  • n.iter Number of iterations if bootstrapped confidence intervals are desired
  • delete Delete items with no variance and issue a warning
  • use Options to pass to the cor function: “everything”, “all.obs”, “complete.obs”, “na.or.complete”, or “pairwise.complete.obs”. The default is “pairwise”
  • warnings By default print a warning and a message that items were reversed. Suppress the message if warnings = FALSE
  • n.obs If using correlation matrices as input, by specify the number of observations, we can find confidence intervals
disgust %>%
  select(moral1:moral7) %>%
  psych::alpha(title = "moral")
## 
## Reliability analysis  moral  
## Call: psych::alpha(x = ., title = "moral")
## 
##   raw_alpha std.alpha G6(smc) average_r S/N    ase mean  sd median_r
##       0.85      0.85    0.84      0.45 5.8 0.0016  3.8 1.3     0.46
## 
##  lower alpha upper     95% confidence boundaries
## 0.85 0.85 0.85 
## 
##  Reliability if an item is dropped:
##        raw_alpha std.alpha G6(smc) average_r S/N alpha se  var.r med.r
## moral1      0.83      0.83    0.81      0.45 4.9   0.0019 0.0035  0.45
## moral2      0.82      0.82    0.80      0.43 4.6   0.0019 0.0033  0.42
## moral3      0.83      0.83    0.81      0.45 5.0   0.0019 0.0040  0.45
## moral4      0.84      0.84    0.82      0.47 5.3   0.0017 0.0023  0.48
## moral5      0.83      0.83    0.81      0.44 4.8   0.0019 0.0038  0.45
## moral6      0.84      0.84    0.82      0.47 5.4   0.0017 0.0033  0.49
## moral7      0.82      0.83    0.81      0.44 4.8   0.0019 0.0043  0.44
## 
##  Item statistics 
##            n raw.r std.r r.cor r.drop mean  sd
## moral1 19668  0.74  0.73  0.67   0.62  3.1 1.9
## moral2 19662  0.77  0.78  0.75   0.68  4.6 1.5
## moral3 19681  0.74  0.73  0.67   0.62  3.2 1.8
## moral4 19656  0.66  0.68  0.60   0.54  4.5 1.5
## moral5 19668  0.76  0.75  0.70   0.64  3.8 1.9
## moral6 19679  0.68  0.67  0.58   0.54  3.8 1.8
## moral7 19665  0.76  0.76  0.70   0.65  3.7 1.7
## 
## Non missing response frequency for each item
##           0    1    2    3    4    5    6 miss
## moral1 0.11 0.13 0.14 0.18 0.18 0.15 0.12 0.02
## moral2 0.03 0.03 0.05 0.09 0.18 0.28 0.34 0.02
## moral3 0.10 0.11 0.13 0.17 0.20 0.17 0.12 0.02
## moral4 0.03 0.03 0.06 0.11 0.19 0.29 0.30 0.02
## moral5 0.07 0.08 0.10 0.15 0.17 0.21 0.23 0.02
## moral6 0.07 0.07 0.10 0.14 0.20 0.22 0.20 0.02
## moral7 0.06 0.08 0.10 0.17 0.21 0.22 0.16 0.02
disgust %>%
  select(sexual1:sexual7) %>%
  psych::alpha(title = "sexual disgust")
## 
## Reliability analysis  sexual disgust  
## Call: psych::alpha(x = ., title = "sexual disgust")
## 
##   raw_alpha std.alpha G6(smc) average_r S/N    ase mean  sd median_r
##       0.81      0.81     0.8      0.38 4.3 0.0021  2.6 1.4      0.4
## 
##  lower alpha upper     95% confidence boundaries
## 0.8 0.81 0.81 
## 
##  Reliability if an item is dropped:
##         raw_alpha std.alpha G6(smc) average_r S/N alpha se  var.r med.r
## sexual1      0.77      0.78    0.76      0.37 3.5   0.0025 0.0073  0.38
## sexual2      0.79      0.79    0.77      0.38 3.7   0.0023 0.0043  0.40
## sexual3      0.77      0.77    0.74      0.36 3.3   0.0025 0.0045  0.38
## sexual4      0.79      0.80    0.77      0.39 3.9   0.0023 0.0073  0.40
## sexual5      0.77      0.78    0.76      0.37 3.5   0.0025 0.0077  0.37
## sexual6      0.80      0.80    0.78      0.40 4.0   0.0022 0.0052  0.40
## sexual7      0.79      0.79    0.77      0.38 3.7   0.0024 0.0078  0.40
## 
##  Item statistics 
##             n raw.r std.r r.cor r.drop mean  sd
## sexual1 19693  0.71  0.72  0.66   0.59  2.4 1.9
## sexual2 19664  0.65  0.67  0.59   0.52  1.4 1.8
## sexual3 19690  0.74  0.75  0.71   0.63  1.6 1.9
## sexual4 19703  0.64  0.64  0.54   0.49  3.0 2.0
## sexual5 19695  0.73  0.72  0.66   0.59  2.7 2.1
## sexual6 19670  0.62  0.62  0.52   0.46  3.9 2.1
## sexual7 19684  0.69  0.67  0.59   0.53  2.9 2.2
## 
## Non missing response frequency for each item
##            0    1    2    3    4    5    6 miss
## sexual1 0.20 0.19 0.16 0.16 0.13 0.09 0.07 0.02
## sexual2 0.47 0.19 0.10 0.09 0.05 0.04 0.05 0.02
## sexual3 0.41 0.20 0.11 0.10 0.07 0.05 0.06 0.02
## sexual4 0.14 0.13 0.14 0.15 0.16 0.14 0.13 0.01
## sexual5 0.20 0.16 0.13 0.14 0.11 0.11 0.14 0.02
## sexual6 0.10 0.08 0.08 0.11 0.12 0.19 0.33 0.02
## sexual7 0.21 0.15 0.11 0.11 0.10 0.12 0.20 0.02
disgust %>%
  select(pathogen1:pathogen7) %>%
  psych::alpha(title = "pathogen disgust")
## 
## Reliability analysis  pathogen disgust  
## Call: psych::alpha(x = ., title = "pathogen disgust")
## 
##   raw_alpha std.alpha G6(smc) average_r S/N    ase mean  sd median_r
##       0.74      0.74    0.72      0.29 2.9 0.0028  3.7 1.1      0.3
## 
##  lower alpha upper     95% confidence boundaries
## 0.73 0.74 0.74 
## 
##  Reliability if an item is dropped:
##           raw_alpha std.alpha G6(smc) average_r S/N alpha se  var.r med.r
## pathogen1      0.71      0.71    0.69      0.29 2.5   0.0032 0.0042  0.26
## pathogen2      0.70      0.71    0.68      0.29 2.5   0.0032 0.0033  0.30
## pathogen3      0.70      0.70    0.67      0.28 2.4   0.0033 0.0028  0.26
## pathogen4      0.71      0.72    0.69      0.30 2.5   0.0032 0.0042  0.30
## pathogen5      0.70      0.70    0.67      0.28 2.4   0.0033 0.0030  0.26
## pathogen6      0.72      0.72    0.70      0.30 2.6   0.0031 0.0042  0.31
## pathogen7      0.71      0.72    0.69      0.30 2.6   0.0031 0.0037  0.30
## 
##  Item statistics 
##               n raw.r std.r r.cor r.drop mean  sd
## pathogen1 19668  0.60  0.63  0.53   0.45  4.4 1.5
## pathogen2 19683  0.64  0.63  0.54   0.46  3.3 1.7
## pathogen3 19687  0.65  0.66  0.58   0.49  3.2 1.6
## pathogen4 19683  0.62  0.62  0.52   0.44  3.7 1.8
## pathogen5 19678  0.64  0.67  0.59   0.50  4.3 1.4
## pathogen6 19655  0.61  0.59  0.48   0.41  3.8 1.9
## pathogen7 19692  0.63  0.61  0.50   0.43  3.5 1.9
## 
## Non missing response frequency for each item
##              0    1    2    3    4    5    6 miss
## pathogen1 0.01 0.04 0.07 0.11 0.22 0.26 0.29 0.02
## pathogen2 0.07 0.12 0.14 0.18 0.22 0.16 0.11 0.02
## pathogen3 0.05 0.14 0.17 0.19 0.24 0.14 0.08 0.02
## pathogen4 0.04 0.11 0.12 0.15 0.21 0.19 0.18 0.02
## pathogen5 0.01 0.04 0.08 0.13 0.24 0.27 0.23 0.02
## pathogen6 0.06 0.10 0.10 0.13 0.18 0.19 0.25 0.02
## pathogen7 0.08 0.12 0.12 0.13 0.17 0.17 0.20 0.02
R alpha psych-r-package cronbach SPSS
Lisa DeBruine
Lisa DeBruine
Professor of Psychology

Lisa DeBruine is a professor of psychology at the University of Glasgow. Her substantive research is on the social perception of faces and kinship. Her meta-science interests include team science (especially the Psychological Science Accelerator), open documentation, data simulation, web-based tools for data collection and stimulus generation, and teaching computational reproducibility.

Related