Background

The state's importance is something that cannot be denied, as is the case with organizations. Basic knowledge of the organization as a particular environment that is structured, characterized, and has specific functions is something that supports the smooth running of the organization. The activity is not only in managing the people who work. Still, it includes the actions of planning, organizing, mobilizing, and controlling the steps taken to determine and achieve targets through the utilization of available resources. This series of activities is called the management process, while the person who leads and manages the management process is called the manager. The role of management in the organization that is managing the organizational structure of control according to the will of the company wants, helping managers in conducting supervision and determining the people needed in the company, will get quick decision making in the company, the organization will be well structured. Proper management must play a role in the organization's situation and conditions, likewise with the organization of the Maritime Security Group (Guskamla) Batam.

In his duties, Guskamla Batam personnel always improve Personnel Professionalism and its quality, both directly and indirectly, but some problems exist, and solutions must be sought. Some of these problems are that the competence of some personnel is not by the part of work that is their responsibility, the training conducted by some staff is still lacking, the character of some personnel to support operations is still not optimal. Based on the description above, the authors plan to research the extent of the determination of these factors, which the authors classify into competence, training, and character factors for loyalty personnel in Batam through personnel professionalism. In addition to examining the determination of the three factors above on loyalty, it also investigates how ersonnel professionalism as an intervening variable, influences loyalty in Guskamla Batam. With this research, it is expected to know how the determination of competence, Training, and Character factors affects Loyalty in Guskamla Batam. Based on this, the next research title of the author is "Determination of Competence,

Exogenous variables (independent variables), namely, Competence (X1), Training (X2), and Character (X3). Intervening variables (variables that mediate or mediate) are Personnel Professionalism (Y), while Endogenous (dependent) variables are Loyalty (Z). The authors take these variables because no one has ever researched Guskamla Batam about the Determination of Competence, Training, and Character of Loyalty through Personnel Professionalism in Guskamla Batam.

Formulation of the Problem

The main things in the formulation of the problem in this study are as follows:

1. Does competence determine the professionalism of personnel?

2. Does the training determine the professionalism of personnel?

3. Does the Character determine the Professionalism of personnel?

4. Does competence determine loyalty?

5. Does training determine loyalty?

6. Does the character determine loyalty?

7. Does personnel professionalism determine loyalty?

Theoretical Framework and Framework for Thinking

Competence

According to Jackson et al. (2010: 205), Competence is a pattern of knowledge, skills, abilities, behaviors, and other measurable characteristics needed by personnel to adequately perform job roles or job functions. According to Suparno (2012: 27), Competence is an adequate ability to carry out a task or possess the required skills and abilities. They are derived from Competence, which means skills, abilities, and authority. Etymologically, Competence is defined as the behavioral dimension of expertise or excellence of a leader or staff who have excellent skills, knowledge, and behavior. The observed person's ability can complete a job or task by predetermined operating standards. But Competence is not just the ability to carry out administrative assignments; Competence will be related to teachings about humans and their behavior, legally humans in carrying out their actions must be by the norms or rules that apply in their lives.

Career management and personnel potential assessment can help the Batam Guskamla organization or agency create career development for personnel to achieve career paths that are in line with their potential. Then the award is given with the achievement of the Competence of staff who will support the implementation of the competency system used by the Batam Guskamla organization as a whole.

Training

According to Prabu Mangkunegara (2013: 44), Training is a short-term educational process that uses systematic and organized procedures where non-managerial employees learn technical knowledge and skills in limited goals. According to Dessler (2015: 284), Training is a process of teaching new or existing employees the necessary skills to carry out their work. Training is one of the efforts to improve the quality of human resources in the world of work. Personnel, either new or who are already working, need to follow Training.

By practicing sincerely, Training is expected; the process is by the plans, programs, and objectives set from the beginning. In other words, Training is a condition that shows how far the project can be achieved. The more plans that can be made, the more effective the activity.

Character

According to Prabu Mangkunegara (2013: 44), Character is a short-term educational process that uses systematic and organized procedures where non-managerial employees teach technical knowledge and skills in limited goals. According to Dessler (2015: 284), Character is a process of teaching new or existing employees the necessary skills they need to carry out their work. Training is one of the efforts to improve the quality of human resources in the world of work. Personnel, either new or who are already working, need to follow Training.

Character is a personality that makes it typical in thinking and acting that is inherent in personnel. The cast consists of three behavioral performances consisting of moral knowledge, feelings based on morals, and behavior based on morals. Good Character consists of knowing where it is good, the desire to do good, and doing good. Character determines whether someone in achieving their ambitions using the right way according to their environment and obey the laws and rules of the group. So, Character is a traitor Character of personnel who can be good or bad based on environmental assessment

Personnel Professionalism

According to Yuwono (2011: 9), Professionalism is a permanent job in a particular field based on specific expertise carried out responsibly, to earn income. A person who has Professionalism is someone who has several items, including:

1. Skill, which means that the person must be indeed an expert in his field, able to do it quickly and precisely, and the results of his work are satisfying.

2. Knowledge means that the person must be able to master, at least insightful about other sciences related to their fields.

3. Attitude, which means not only smart but must have ethics applied in the area.

Loyalty

According to Griffin (2010: 04), Loyalty is a consumer being said to be loyal or loyal if the consumer shows a buying behavior regularly, or there is a condition where it requires the consumer to buy at least twice in a specified time interval. Loyalty is one thing that cannot be purchased with money. Reliability can only be obtained, but cannot be bought. Gaining Loyalty from the personnel is not an easy job to do. Inversely proportional to the difficulty of getting it, eliminating the Loyalty of staff becomes a straightforward thing.

Full Model Competence, Training, Character, Personnel Professionalism, and Loyalty.

With Personnel Professionalism as a Mediator between Competence,

Training and Character for Loyalty

Research Model

The hypotheses tested are:

From Figure 1.6 above, the structural equation model can be made as follows:

H1: Y = ?y.x1 X1 + e1,? the direct effect (Direct Effects) X1 on Y,

H2: Y = ?y.x2 X2 + e1, ? the direct effect (Direct Effects) X2 on Y,

H3: Y = ?y.x3 X3 + e1, ? the direct effect (Direct Effects) X3 on Y,

H4: Z = ?z.x1 X1 + e2, ? the direct effect (Direct Effects) X1 on Z,

H5: Z = ?z.x2 X2 + e2, ? direct effect (Direct Effects) X2 on Z

H6: Z = ?z.x3 X3 + e2, ? direct effect (Direct Effects) X3 on Z

H7: Z = ?zy Y1 + e2, ? direct effect Y on Z

Population and Sample

The population in this research is the Batam Guskamla personnel. According to Romanian (2010: 140), the community is a group of people, events, things that have specific characteristics. Population representation through samples taken. All members of the population were sampled. This study used a saturated sample taken from the Census Technique using proportional random sampling, namely personnel in the Guskamla Batam organizational units. The five staff consisting of 120 personnel, were made into the population, the sample in this study. The data has been confirmed with Spurs Koarmada I based on the recapitulation of data on the number of personnel as of December 2019.

Data Analysis Technique

Data obtained in this study were analyzed. Data analysis techniques in this study using AMOS. There are two stages to the group in analyzing SEM:

(1) Measurement Model Analysis (Outer Model),

(2) Structural Model Analysis (Inner Model).

Analysis of the measurement model (Outer Model) aims to evaluate the construct variables being studied, namely the validity (accuracy) and reliability (reliability) of a variable, including:

(1) Internal consistency / composite reliability,

(2) Convergent Validity / Average Variance Extracted / AVE, and

(3) Discriminant Validity (Hair, Hult, Ringle, & Sarstedt, 2014).

Influence Analysis with SEM

By the purpose of the study to determine the effect of Competence, Training, Character, Personnel Professionalism and Loyalty and hypotheses, the data analysis is performed using Structural Equation Modeling (SEM) which is a set of statistical techniques that allows testing a series of relatively complex relationships simultaneously (Ferdinand, 2002; Solimun, 2004). Referring to the picture above, we can see the number of indicators of each latent variable Competence, Training, Character, Personnel Professionalism, and Loyalty. Respondents' answer scores were processed with Full Model Structural Equation Modeling (SEM) statistics using AMOS for windows software version 24.0 to obtain graphical displays and Regression Weights results.

From the results of data processing, Full Model Structural Equation Modeling (SEM) was analyzed. For analysis, the first image displayed is the result of processing Structural Equation Modeling (SEM).

Regression Coefficient Model Variable Competence, Training, Character, Personnel Professionalism, and Loyalty

Full Model SEM

Based on the picture above, it can be analyzed the measurement model with lambda parameters, structural model analysis, Goodness of fit and determination analysis for the influence of Competence, Training, Character, Personnel Professionalism, and Loyalty.

Testing Analysis of Measurement Models with Lamda Parameters

The parameter testing carried out was Lamda parameter testing. This test is intended to determine the validity of each research indicator for testing the lambda parameter used standardized estimate (regression weight) value in the form of loading factor.

If the value of standardized estimate (regression weight)> 0.50, the value of CR> table = 2,000 and Probability <= 0.05, then the loading factor parameter lambda, the indicator is declared significant (Ferdinand, Agusty, 2002: 97). This means the indicator is valid. To test these parameters, the following table displays the loading factor/lambda CR, Probability (P).

Analysis of Structural Equation Models

Model testing is done using regression coefficients for the variables Competence (X1), Training (X2), Character (X3), Personnel Professionalism (Y), and Loyalty (Z) through the output table from the submenu view/set. Based on the calculation of the regression coefficient (regression weight), an output table is made, as presented in Table 1.18, as follows.

Table 1. Standardized Direct Effects (Group number 1 - Default model) Competence (COM), Training (TRA), Character (CHA), Personnel Professionalism (PRO) and Loyalty (LOY)

Table 2. Regression Weight Competence (COM), Training (TRA), Character (CHA), Personnel Professionalism (PRO) and Loyalty (LOY)

Table 3. Standardized Regression Weight Competence (COM), Training (TRA), Character (CHA), Personnel Professionalism (PRO) and Loyalty (LOY)

Based on the three Tables above, it Appears

The effect of the Competency latent variable (COM) on the latent variable Personnel Professionalism (PRO) has a standardized estimate (regression weight) of 0.080 with Cr (Critical ratio = identical to the t-count value) of 0.771 at probability = 0.441. CR value of 0.771 <2.00 and Probability = 0.441> 0.05 indicates that the influence of the Competent latent variable (COM) on the latent variable Professional Personnel (PRO) is positively insignificant. Competence is a person's ability that is developed over a long time by leaders and members of the organization that are socialized and taught to new personnel and applied in organizational activities so that it does not influence Personnel Professionalism because Personnel Professionalism is different from other parts, in an organization that carries out human resource management processes. This has caused the latent variable Competence (COM) to the latent variable Professional Personnel (PRO), a positive but not significant effect.

The effect of latent training (TRA) variables on the latent variable Professional Personnel (PRO) has a standardized estimate (regression weight) of 0.575 with Cr (Critical ratio = identical to the t-count value) of 3.888 on Probability = ***. CR value of 3.888> 2,000 and Probability = *** <0.05 indicates that the effect of the latent variable Training (TRA) on the latent variable Professional Personnel (PRO) is significantly positive. Conformity with the theory that, comprehensively, Training is defined as an activity found or developed by a particular group to learn to overcome personal and group problems from external adaptation and internal integration. So the latent variable Training (TRA) of the latent variable Professional Personnel (PRO), a significant positive effect.

The effect of the Character latent variable (CHA) on the Professional Personnel latent variable (PRO) has a standardized estimate (regression weight) of 0.498 with Cr (Critical ratio = identical to the t-count value) of 3.322 at Probability = ***. CR value = 3.322> 2,000 and Probability = *** <0.05 shows that the influence of the Character latent variable (CHA) on the Professional Personnel (PRO) latent variable is significantly positive. Character refers to a series of attitudes (attitudes), behavior (behavior), motivation (motivation), and skills (skills)—WhileProfessionalismare the properties (abilities, skills, ways of implementing things, etc.) as is reasonable. Someone who has a spirit of Professionalism always pushes himself to realize professional work. So that the latent variable Character (CHA) on the latent variable Professional Personnel (PRO), significant positive effect.

The effect of the Competence latent variable (COM) on the Loyalty latent variable (LOY) has a standardized estimate (regression weight) of 0.383 with Cr (Critical ratio = identical to the t-count value) of 1.840 at probability = 0.066. CR value 1,840 <2,000 and Probability = 0.066> 0.05 shows that the influence of the Competency latent variable (COM) on the Loyalty latent variable (LOY) is not significant, because Competence does not significantly affect Loyalty, even though it is within an organization in conducting management processes human Resources. This is what causes the Competency latent variable (COM) to the Loyalty latent variable (LOY), a positive but not significant effect.

The effect of the Training latent variable (TRA) on the Loyalty latent variable (LOY) has a standardized estimate (regression weight) of 0.862 with Cr (Critical ratio = identical to the t-count value) of 1.948 at probability = 0.051. CR value of 1.948 <2,000 and Probability = 0.051> 0.05 indicates that the effect of the Training latent variable (TRA) on the Loyalty latent variable (LOY) is positively insignificant. An organization recommends many pieces of Training. It is hoped that participants will share their feelings with these values and views on Training (TRA), which are the norms, organizational habits, and so on, developed shortly by leaders and members of the organization. They are socialized and taught to personnel. Training will have a positive impact on the ability and expertise of staff to be able to guide personnel towards higher career paths. Therefore, Training does not significantly affect Loyalty, which is in one organization in carrying out human resource management processes. This is what causes the Competency latent variable (COM) to the Loyalty latent variable (LOY), a positive but not significant effect.

The effect of the Character latent variable (CHA) on the Loyalty latent variable (LOY) has a standardized estimate (regression weight) of 0.860 with a Cr (Critical ratio = identical to the t-count value) of 1,990 at probability = 0.047. CR value of 1.990 <2,000 and Probability = 0.047 <0.05 indicates that the influence of the Character latent variable (CHA) on the Loyalty latent variable (LOY) is positively insignificant. Character is innate, heart, soul, personality, Character, behavior, personality, Character, Character, temperament, and Character. Whereas Loyalty is an attitude that someone to obey the rules. These two things are related but different so that the latent variable Character to the Loyalty latent variable is positively insignificant.

The effect of the Professional Personnel (PRO) latent variable on the Loyalty latent variable (LOY) has a standardized estimate (regression weight) of - 0.919 with Cr (Critical ratio = identical to the t-count value) of -1.266 at probability = 0.206. CR value -1.266 <2,000 and Probability = 0.206> 0.05 shows that the effect of the latent variable Professional Personnel (PRO) on the hidden variable Loyalty (LOY) is positively insignificant. Professional Personnel (PRO) itself is the ability to act professionally, as much as possible, trying to be a good role model, discover and realize the enormous potential possessed by each individual, and about what can be done with several things and for reputation. In contrast, Loyalty is a person's obedience and Loyalty.

Taking into account the standardized estimate for the Competence (X1), Training (X2), Character (X3), Professional Personnel (Y), and Loyalty (Z) variables, the structural equation model can be made as follows:

H1: Y = ? y.x1 X1 + e1 = 0,080 X1 + e1

H2: Y = ? y.x2 X2 + e1 = 0.575 X2 + e1

H3: Y = ? y.x3 X3 + e1 = 0,498 X3 + e1

H4: Z = ? z.x1 X1 + e2 = 0.383 X1 + e2

H5: Z = ?z.x2 X2 + e2 = 0.862 X2 + e2

H6: Z = ? z.x3 X3 + e2 = 0.860 X3 + e2

H7: Z = ? zy Y + e2 = - 0,919 Y + e2

Analysis of Goodness of Fit

Based on test criteria, Chi-square (c²), Relative Chi-square (c²/ df), RMSEA, GFI, AGFI, TLI, and CFI above and the Goodness of Fit value of Amos for windows processing version 24.0, then the following table can be made.

Table 4. The Goodness of Fit Evaluation

*) Meets Goodness of fit+) Marginal

Noting the cut-of-value and Goodness of fit results of the model in Table 1.21 above, one good and one marginal of the eight criteria are used. The right measures are Chi-square, and minimal rules are RMSEA. Because only one is correct, and one is negligible from the eight standards required, the above model can be stated as a wrong model (Solimun, 2002: 80 and Solimun, 2004: 71).

Analyze the Measurement Model with Determination

The following is an analysis of the Measurement Model with the coefficient of Determination of Competence (COM), Training (TRA) and Character (CHA) on Personnel Professionalism (PRO) and the influence of Competence (COM), Training (TRA) and Character (CHA) and Professional Professionalism (PRO) towards Loyalty (LOY). Analysis of the measurement model with Determination is used to determine the contribution of exogenous variables to endogenous variables. For this analysis, Square, Multiple Correlation, is used. The amount of Square Multiple Correlation can be seen in the following table.

They squared Multiple Correlation whose value is for Personnel Professionalism (PRO) = 1,107 for Loyalty (LOY) = 0.789, as shown in Table 1.22 above. According to Ferdinand, 2002: 114, Square Multiple Correlation values ??for the Personnel Professionalism (PRO) R2 = 1,107 are identical to R2 in SPSS of 1,107. The amount of Determination is the Squared Multiple Correlation value for the Personnel Professionalism variable times 100% = 1,107 x 100% = 110, 7%. Thus, the change in Personnel Professionalism (PRO) is influenced by Competence, Training, and Character by 110.7%.

For Loyalty (LOY) R2 = 0, 789 then the amount of Determination = 0, 789 x 100% = 78.9%. Thus, it can be stated that Loyalty changes are influenced by Competence, Training, Character, and Personnel Professionalism of 78.9%.

From the description above it can be stated that there are seven variables, 2 of which have a significant positive effect and 5 of the variables which have a positive impact that is not significant as stated below:

a) Hypothesis 1 (H1): Competence influences Personnel Professionalism, but it is not significant

b) Hypothesis 2 (H2): Training affects Personnel Professionalism with its validity.

c) Hypothesis 3 (H3): Character influences Personnel Professionalism is proven right.

d) Hypothesis 4 (H4): Competence influences the truth of Loyalty, but it is not significant

e) Hypothesis 5 (H5): Training has an effect on Loyalty which is proven to be correct but not significant

f) Hypothesis 6 (H6): Character influences proven truth but not significant Loyalty.

g) Hypothesis 7 (H7): Personnel Professionalism affects Loyalty, but the truth is not significant.