I am trying to compute internal consistency of a questinnaire with 37 items out of which 8 items have reverse scoring. While entering the data I have already entered the reverse scored data. And this data is being called by R to compute internal consistency with the help of alpha() function. How d I tell the code that these were the specific items which were reverse coded (in a 5-point likert scale, it would mean that if the participant marked 1, it is entered as 5 in the data) for the calculation of alpha.
I tried the following code and it gave me few warnings.
#reading table
raw_data <- read.table("raw_data_117.csv", sep = "," , header=T)
#Surgency
Surgency <- select(raw_data, X1, X4, X7, X10, X13.R, X16, X19.R, X22.R, X25, X28, X31.R, X34.R)
alpha(Surgency)
Following was the output:
Some items ( X19.R X22.R X31.R X34.R ) were negatively correlated with the total scale and
probably should be reversed.
To do this, run the function again with the 'check.keys=TRUE' option
Reliability analysis
Call: alpha(x = Surgency)
raw_alpha std.alpha G6(smc) average_r S/N ase mean sd median_r
0.61 0.62 0.72 0.12 1.6 0.054 4.6 0.91 0.12
95% confidence boundaries
lower alpha upper
Feldt 0.49 0.61 0.70
Duhachek 0.50 0.61 0.71
Reliability if an item is dropped:
raw_alpha std.alpha G6(smc) average_r S/N alpha se var.r med.r
X1 0.58 0.59 0.68 0.116 1.4 0.058 0.035 0.128
X4 0.53 0.55 0.64 0.099 1.2 0.065 0.030 0.105
X7 0.59 0.60 0.70 0.119 1.5 0.058 0.035 0.113
X10 0.59 0.60 0.70 0.122 1.5 0.057 0.041 0.094
X13.R 0.58 0.60 0.69 0.119 1.5 0.059 0.041 0.113
X16 0.58 0.60 0.68 0.118 1.5 0.058 0.034 0.113
X19.R 0.58 0.60 0.69 0.118 1.5 0.059 0.038 0.128
X22.R 0.62 0.63 0.71 0.132 1.7 0.052 0.034 0.148
X25 0.57 0.58 0.68 0.111 1.4 0.059 0.039 0.113
X28 0.59 0.60 0.70 0.122 1.5 0.057 0.037 0.113
X31.R 0.62 0.64 0.72 0.139 1.8 0.052 0.036 0.148
X34.R 0.60 0.62 0.71 0.129 1.6 0.055 0.039 0.148
Item statistics
n raw.r std.r r.cor r.drop mean sd
X1 110 0.47 0.49 0.45 0.293 5.3 1.7
X4 111 0.64 0.66 0.69 0.501 4.9 2.1
X7 112 0.41 0.45 0.38 0.260 4.8 1.9
X10 112 0.44 0.43 0.34 0.252 4.9 2.0
X13.R 109 0.50 0.46 0.38 0.296 4.7 2.3
X16 108 0.47 0.46 0.43 0.283 4.7 2.2
X19.R 113 0.47 0.46 0.39 0.314 4.2 2.0
X22.R 109 0.36 0.32 0.22 0.135 3.9 2.5
X25 117 0.51 0.54 0.49 0.366 6.2 1.4
X28 103 0.44 0.43 0.34 0.250 3.6 2.3
X31.R 108 0.29 0.24 0.12 0.068 4.6 2.0
X34.R 113 0.36 0.35 0.25 0.189 3.0 1.9
Non missing response frequency for each item
1 2 3 4 5 6 7 miss
X1 0.06 0.02 0.06 0.08 0.21 0.29 0.27 0.06
X4 0.15 0.04 0.08 0.05 0.15 0.25 0.27 0.05
X7 0.12 0.02 0.09 0.14 0.24 0.16 0.23 0.04
X10 0.12 0.03 0.12 0.10 0.12 0.23 0.28 0.04
X13.R 0.12 0.12 0.10 0.13 0.06 0.06 0.41 0.07
X16 0.15 0.08 0.06 0.06 0.15 0.20 0.29 0.08
X19.R 0.12 0.08 0.28 0.05 0.18 0.11 0.19 0.03
X22.R 0.24 0.19 0.10 0.05 0.04 0.07 0.31 0.07
X25 0.03 0.01 0.02 0.07 0.07 0.18 0.63 0.00
X28 0.33 0.10 0.08 0.09 0.14 0.12 0.16 0.12
X31.R 0.09 0.05 0.21 0.14 0.07 0.15 0.29 0.08
X34.R 0.27 0.22 0.20 0.10 0.07 0.03 0.12 0.03
Warning message:
In alpha(Surgency) :
Some items were negatively correlated with the total scale and probably
should be reversed.
To do this, run the function again with the 'check.keys=TRUE' option
>
I have already entered the reverse coded data and need not do it through codes. Please let me know whether what I have done is correct or not. Items with .R were negative items and were reverse coded in the table raw_data prior to passing it to the function alpha().