leadership and motivation in hospitality

cent) from the Work Values (WV) construct. Neither of these violates the ‘greater
than one third’ guideline for this research.

Following the removal of these cases, the characteristics of the missing values in
the data set were re-evaluated and can be summarised as follows:

 17 (37 per cent) of the 46 variables have no missing values;
 177 (83 per cent) of the 213 cases have no missing values; and
 99.5 per cent of data points are complete.

Missing data distribution by variable is summarised in Table 6 - 3 and their
distribution by case in Table 6 - 4.

The distribution of missing data by variable continues to more than adequately
satisfy rule of thumb (B) which states that variables with 15 per cent missing data
are candidates for deletion. Table 6 - 3 shows that missing values account for no
more than 2.8 per cent in the affected variables.

Variable/s
(n = 213)

Number of missing

observations

Percentage of
missing
observationsa
JS5 6 2.8
ME1, SS4 4 1.9
EM3, EM7, ME6 3 1.4
EM6, JS2, ME3, ME7, SS1, WV5 2 0.9
AC1, AC2, AC4, DSB4, EM9, JS3, JS4,
JS6, SS3, WV1, WV4, WV6, WV7 1 0.5^
a – based on 46 item statement variables

Table 6-3 Distribution of missing values by variable

Regarding the level of missing values per case (see Table 6 - 4 ), only one case
(with 5 missing values) is flagged for attention. Attention is warranted because
greater than 10 per cent of all data points are missing; specifically, 5 missing
values = 10.9 per cent of all data from the 46 item statements.

The five missing values are found across four different latent constructs as
follows:

 1 in Motivational Leadership (ML4);
 1 in Work Meaning (ME7);
 1 in Employee Empowerment (EM7); and
 2 in Job Satisfaction (JS2 and JS5).