According to the coding of the two members of the research group, "familiarity with the use of office software" was not coded, indicating that this competency element was not reflected in the text. Therefore, this competency element was removed and the remaining 26 competency elements were analyzed.
2.1 Analysis of Interview Length
The length of the interview can be reflected by the time consumption of the interview or the number of words in the text, and both variables are normally distributed. The purpose to test the difference of interview length between different performance groups, is to exclude the impact of interview length on the competency analysis of the two groups.The average interview time of the high achievement group was 39 minutes, and 43.3 minutes of the average achievement group; the average number of words in the interview text of the high achievement group was 3295.25 words, and 2993 words of the average achievement group. Two independent sample t-tests were conducted on the interview length of different performance groups, and there was no significant difference between groups (p values were 0.549 and 0.594 respectively, both greater than 0.05), indicating that the competency differences of different performance groups were not affected by the interview length, as shown in Table 2.
Table 2 Interview Length of Different Performance Groups

At the same time, by correlation analysis, it could be seen that when the three variables of average score, highest score and total frequency were adopted, only the application ability of laws and regulations was positively correlated with the number of words in the interview text twice (p values are 0.013 and 0.021 respectively, all less than 0.05). See Table 3. Other competency elements were not affected by the number of words in the interview text, It showed that the three variables had good stability.
Table 3 Correlation Analysis Between the Number of Words of Interview Text and Statistical Indicators of Competency
Competency element
|
Highest score to the number of words
|
Average score to the number of words
|
Total frequency to the number of words
|
r
|
p
|
r
|
p
|
r
|
p
|
Professional knowledge
|
0.18
|
0.699
|
0.317
|
0.489
|
0.661
|
0.106
|
Legal knowledge
|
0.167
|
0.72
|
0.237
|
0.608
|
0.543
|
0.208
|
Ability to produce law enforcement instruments
|
-0.017
|
0.971
|
-0.017
|
0.971
|
0.144
|
0.758
|
Application ability of laws and regulations
|
0.829*
|
0.021
|
0.862*
|
0.013
|
0.495
|
0.259
|
Teamwork ability
|
0.145
|
0.756
|
0.174
|
0.708
|
-0.205
|
0.659
|
Adaptability
|
0.161
|
0.73
|
0.121
|
0.796
|
0.412
|
0.358
|
Investigation and evidence collection ability
|
0.056
|
0.905
|
0.323
|
0.48
|
0.738
|
0.058
|
Communication and coordination skills
|
0.178
|
0.703
|
0.165
|
0.723
|
0.085
|
0.856
|
On site control capability
|
0.189
|
0.685
|
0.272
|
0.554
|
-0.145
|
0.756
|
Judgment
|
0.125
|
0.79
|
0.125
|
0.79
|
0.485
|
0.27
|
Inquiry skills
|
0.274
|
0.552
|
0.141
|
0.763
|
-0.368
|
0.417
|
Analytical ability
|
0.582
|
0.171
|
0.536
|
0.215
|
0.751
|
0.051
|
On site inspection skills
|
0.29
|
0.528
|
0.29
|
0.528
|
0.29
|
0.528
|
Scientific research ability
|
-0.247
|
0.593
|
-0.259
|
0.575
|
-0.301
|
0.512
|
Learning ability
|
-0.145
|
0.756
|
-0.062
|
0.896
|
0.411
|
0.36
|
Honesty and integrity
|
-0.126
|
0.787
|
-0.126
|
0.787
|
-0.126
|
0.787
|
Legal awareness
|
0.371
|
0.412
|
0.371
|
0.412
|
0.054
|
0.908
|
Justice awareness
|
-0.126
|
0.787
|
-0.126
|
0.787
|
-0.126
|
0.787
|
Sense of responsibility
|
-0.204
|
0.661
|
-0.204
|
0.661
|
0.494
|
0.259
|
Initiative
|
-0.172
|
0.712
|
-0.291
|
0.526
|
0.01
|
0.983
|
Service awareness
|
0.112
|
0.811
|
0.112
|
0.811
|
0.186
|
0.689
|
Innovation
|
-0.403
|
0.37
|
-0.403
|
0.37
|
-0.373
|
0.41
|
Physical quality
|
-0.643
|
0.119
|
-0.618
|
0.139
|
-0.672
|
0.099
|
Achievement tendency
|
-0.207
|
0.657
|
-0.191
|
0.681
|
-0.243
|
0.6
|
Tenacity and patience
|
0.185
|
0.692
|
0.185
|
0.692
|
-0.456
|
0.304
|
Career love
|
0.308
|
0.502
|
0.285
|
0.535
|
0.215
|
0.643
|
2.2 Coding Reliability Test
We used Cohen's kappa coefficient to test whether the coder's codes were consistent. As shown in Table 4, except for one coefficient value of 0.583, the others were above 0.60. According to Cohen's kappa coefficient consistency judgment standard, the coefficient value is between 0.61-0.80, which means that they have strong consistency. It could be seen that the coding consistency of the two coders was relatively high.
Table 4 Consistency of Two Coders
Text number
|
Cohen's Kappa coefficient
for Competency coding
|
Cohen's Kappa coefficient
For grade score
|
1
|
0.819
|
0.705
|
2
|
0.783
|
0.583
|
3
|
0.777
|
0.645
|
4
|
0.834
|
0.648
|
5
|
0.876
|
0.762
|
6
|
0.850
|
0.700
|
7
|
0.964
|
0.895
|
2.3 Analysis of Competency Difference
2.3.1 Difference Test of Mean Grade Score
The average scores of 26 competency elements in the high performance group and the average performance group were tested by normal distribution. Among them, 17 elements such as “legal knowledge”, “application ability of laws and regulations” and so on did not conform to the normal distribution (Kolmogorov Smirnov test, p < 0.05), and other 9 elements conformed to the normal distribution (Kolmogorov Smirnov test, P > 0.05). Two independent sample t-tests were conducted between the two groups of competency elements for the average scores conforming to the normal distribution. There were significant differences in the ”investigation and evidence collection ability” ,” communication and coordination ability”, “on-site control ability” and “teamwork ability” (p values were 0.046, 0.044, 0.000 and 0.030, all less than 0.05). Kruskal Wallis test was conducted for the average scores of competency characteristics that did not conform to the normal distribution. There were significant differences in "career love" among different performance groups (P value was 0.022, less than 0.05). See Table 5 for specific results.
Table 5 Test for Difference of Average Score between the High Performance Group and the Average performance Group

2.3.2 Difference Test of Highest Grade Score
Normal distribution test was conducted for the highest grade scores of 26 competency elements in the high performance group and the average performance group. Among them, 16 items such as “legal knowledge”, “the application ability of laws and regulations” and so on did not conform to the normal distribution (Kolmogorov Smirnov test, P < 0.05), and the other 10 items conformed to the normal distribution (Kolmogorov Smirnov test, P > 0.05). Two independent sample t-tests were conducted between the two groups of data conforming to normal distribution. There were significant differences in “investigation and evidence collection ability”, “communication and coordination ability”, “on-site control ability”, “teamwork ability” and “career love” between different performance groups (p values were 0.004, 0.047, 0.000, 0.045 and 0.002, all less than 0.05), Kruskal Wallis test was conducted for the highest score of competency elements that did not conform to the normal distribution. There was no significant difference between the high performance group and the average performance group. See Table 6 for specific results.
Table 6 Difference Test of The Highest Grade Score between the High Performance Group and the Average performance Group

2.3.3 Difference Test of Total Frequency
The total frequency of 26 competency elements in the high performance group and the average performance group was tested by normal distribution. Among them, 17 elements such as “legal knowledge”, “the ability to make law enforcement documents” and so on did not conform to the normal distribution (Kolmogorov Smirnov test, P < 0.05), and the other 9 elements conformed to the normal distribution (Kolmogorov Smirnov test, P > 0.05). Similarly, two independent sample t-tests were conducted on the total frequency of 9 competency elements conforming to the normal distribution. There were significant differences in “investigation and evidence collection ability”, “on-site control ability”, “legal awareness” and “career love” between different performance groups (P values were 0.039, 0.007, 0.018 and 0.007, all less than 0.05), Kruskal Wallis test was conducted on the total frequency of 9 competency elements that did not conform to the normal distribution. There was no significant difference between the high performance group and the average performance group. See Table 7 for specific results.
Table 7 Test for Difference of Total Frequency Between the High Performance Group and the Average performance Group

2.3.4 Analysis of Competency of Excellent Front-line Health Inspector
Based on the difference test results of the above three variables, for different performance groups, there were 5 competency elements with statistical differences in average grade scores (P < 0.05), there were 5 competency elements with statistical difference in the highest grade score (P < 0.05), there were 4 competency elements with statistical differences in the total frequency (P < 0.05), and the average value of the three variables for the high performance group was greater than that of the average performance group. It is concluded that the prominent competency elements of excellent front-line health inspector can include the following 6 items: “investigation and evidence collection ability”, “on-site control ability”, “legal awareness”, “career love”, “communication and coordination ability” and “teamwork ability”, and the high performance group had higher value on all the six competency elements.
2.3.5 Results of Expert Consultation
26 competency elements were compiled into a questionnaire, and 20 experts selected the prominent competency elements that the excellent front-line health inspector should have independently. The selection frequency of "investigation and evidence collection ability”, “on-site control ability”, “legal awareness”, “career love”, “communication and coordination ability” and ”teamwork ability" was all more than 80%, ranking in the top 10. The expert consultation results were consistent with the results of BEI. It can be explained to a certain extent that the prominent competency elements of excellent front-line health inspectors extracted by BEI are scientific and reasonable.