This research was developed at Hotel X, a recognized and prestigious company of the Varadero Tourist Center. It was carried out during the months of December and January 2008 and 2009, respectively.
A search and bibliographic review of documents, brochures, books and websites on the subject to be investigated, outside and within the entity, was carried out, mainly using the academic bibliography and official documents of Hotel X. Surveys, questionnaires and interviews were applied to workers and managers of the center.
The general objective of the research was to establish the ranking of the five restaurants in hotel X. For this, various statistical mathematical tools were used for decision-making, including: brainstorming, stratified sampling with proportional allocation, distribution table of frequencies, frequency histogram, Pareto chart, consensus coefficient, coefficient of expertise or competence, weighting using Hierarchical Analytical Process and Modified Füller's Triangle, measures of central tendency, in this case the mode and median, the Ordered Preference method (TOPSIS), and the Kruskal-Wallis test. Some of them were processed through software such as: SPSS, version 11.5, DECISOFT version 2.0 and Statgraphics Plus version 15.1.
The final results correspond to the estimates made initially, and in general it can be said that in the installation, the investigation could be carried out successfully.
METHODOLOGICAL PROCEDURE
To make the ranking of the five restaurants in hotel X, it is necessary to design a methodological procedure that allows, through a sequence of steps, to reach a final result. Described below:
- Define the overall goal. Identify evaluation criteria.
To obtain the criteria, brainstorming is carried out. For it:
- The population or universe of employees made up of the workers of each of the restaurants is determined. The sample size is calculated for a finite population. A stratified sampling with proportional allocation is carried out. The employees to be surveyed are selected using the systematic method. written to obtain items that measure the performance of restaurants The initial number of attributes is refined with the Pareto diagram The resulting criteria are represented and organized in a frequency distribution table
- Choose the decision alternatives.
In the hotel the five existing restaurants are studied.
- Analyze possible experts within the population or universe of workers. Initially, those whose workforce recognizes their experience and knowledge are taken into account
- They self-apply the questionnaire to calculate the coefficient of expertise or competence Those who meet the requirement K ≥ 0.8 are selected as experts
- Refine the identified criteria.
The consensus coefficient calculation method is applied among the experts.
- Weight the criteria.
The Hierarchical Analytical Process (PAJ) method is used:
- A criteria-criteria matrix is drawn up. The measurement scale from 1 to 9 proposed by the authors of the PAJ is taken. The opinion of a single expert is assumed, the one with the greatest knowledge and experience in hotel restoration.
- Apply the corresponding method to rank the decision alternatives.
We work with the Ordered Preference method (TOPSIS):
- A matrix of criteria-alternatives is made The Likert- type measurement scale is taken The opinion of the selected experts is assumed A single matrix is made by applying the measure of central tendency: median, to the set of matrices answered by the experts The non Existence of significant differences between the criteria or opinions of the experts The TOPSIS method is developed The order of descending preference of the alternatives is obtained
- Make recommendations and propose improvement strategies.
In the hotel, based on the order obtained, strong and weak points are identified in each of the alternatives (restaurants), and improvement strategies are outlined.
VALIDATION OF THE METHODOLOGICAL PROCEDURE
A global goal was defined to rank the five restaurants in the hotel as shown below:
To identify the evaluation criteria, it is determined that the population or universe that includes the total number of workers in each of the restaurants is:
Subsequently, the sample size for a finite population is calculated, using the equation:
n = Z 2 * P * Q * N
e 2 * N + Z 2 * P *
Q
Where:
e: error admitted by the collective of authors
Z: percentile of the normal distribution corresponding to the value of e
P: probability of success
Q: probability of failure
N: population size
e = 0.10 Z = 2.58 P = 0.5 Q = 0.5 N = 103
N = 64
The sample size obtained is 64 employees.
Then a stratified sampling with proportional allocation is carried out as follows:
n = 64 = 0.6214
N 103
In order to brainstorm in writing, the Human Resources department was asked to list the number of workers per Restaurants, thus being able to select who would be surveyed for the brainstorming. In this way, brainstorming was carried out in order to obtain items that would measure the performance of the Restaurants. The collected opinions were tabulated in a table shown below:
ITEMS or
Setting 12
Quality of supply 23
Sufficient amount of A + B 5
Comfort 16
Lineup 5
Physical presence of
10
dependents
Professionalism 48
Quality of service 46
Hygiene 16
Variety of dishes 9
Optimal amount of labor 6
Good working conditions 10
Good address 7
To discriminate the quantity of attributes obtained, the following Pareto diagram was drawn:
Pareto chart
Brainstorming
Criteria issued
According to the information shown in the graph, the most notable and vital attributes that measure the performance of Restaurants are:
Frequency distribution table
To refine the number of attributes as much as possible, the consensus coefficient is calculated for which the opinion of several experts is necessary. To do this, it begins by analyzing those workers whose workforce recognizes that they have experience and knowledge in the hotel's restaurant activity. These potential experts were given a questionnaire to answer that would allow them to find the coefficient of expertise or competence of each one. The results are shown below:
Possible Expert 1:
List of characteristics Priority Voting
K c = 0.181 0.054 + 0.100 + 0.113 + 0.122 + 0.145 + 0.018 K c = 0.733
K a = 0.66
Coefficient of expertise: K = K c + K a
two
K = 0.733 + 0.66 K = 0.697
two
If 0.8 ≤ K ≤ 1 then the person can be considered an expert. In this case, possible expert 1 is not an expert (K <0.8).
Possible Expert 2:
K c = 1
K a = 0.76
K = 0.88
Prospective Expert 2 can be considered an Expert.
Possible Expert 3:
K c = 0.824
K a = 0.82
K = 0.822
Potential Expert 3 is considered an Expert.
Possible Expert 4:
K c = 0.982
K a = 0.92
K = 0.95
Possible expert 4 is considered as an expert.
Possible Expert 5:
K c = 1
K a = 0.90
K = 0.95
Potential expert 5 is considered an expert.
Possible Expert 6:
K c = 0.602
K a = 0.72
K = 0.661
Potential expert 6 cannot be considered an expert.
Possible Expert 7:
K c = 0.637
K a = 0.54
K = 0.589
Potential expert 7 will not be considered an expert.
Possible Expert 8:
K c = 1
K a = 0.78
K = 0.89
The potential expert 8 is considered an expert.
The previous calculation of each coefficient of expertise (K) corresponding to the eight possible experts on the subject of hotel restoration, resulted in five real experts according to the rule that states: “When K ≥ 0.8 then the person is considered expert ”.
Next, the consensus coefficient will be found using the five experts who will make their judgment about the eight criteria or attributes that resulted from the Pareto diagram:
- P: professionalism CS: quality of service CO: quality of the offer H: hygiene C: comfort A: atmosphere PF: physical presence of employees CT: good working conditions
If the degree of acceptance of the attributes or criteria is between 0.80 and 1 (0.80 ≤ G c ≤ 1) then the criterion is accepted to carry out the study.
G c = (1- V n / V t) * 100%
G c1 = (1-0 / 5) 100% G c2 = (1-0 / 5) 100% G c3 = (1-0 / 5) 100% G c4 = (10/5) 100%
G c1 = 100% G c2 = 100% G c3 = 100% G c4 = 100%
G c5 = (1-3 / 5) 100% G c6 = (1-4 / 5) 100% G c7 = (1-1 / 5) 100% G c8 = (12/5) 100%
G c5 = 40% G c6 = 20% G c7 = 80% G c8 =
60%
According to the opinion of the experts and fulfilling the aforementioned condition for the coefficient, the study will continue with the following criteria:
- Professionalism Quality of service Quality of supply Hygiene Physical presence of employees
The investigation will continue evaluating the five restaurants (or alternatives) according to five criteria (or attributes) using five experts (or decision makers). The theory that it raises is fulfilled: "The number of experts cannot exceed the number of criteria."
To determine the weight or weighting of each of the five criteria, the Hierarchical Analytical Process method is used, using the information of the expert with the most knowledge on the subject of hotel catering. The scale with values from 1 to 9 proposed by the method and made known to the expert, serves to complete the following Saaty matrix:
For the five criteria in terms of the overall goal:
- Sum the values in each column of the weighted criteria matrix Build the normalized weighted criteria matrix Convert the normalized weighted criteria matrix to decimal form and average the elements in each row
P CS CO H PF W
P 0.043 0.006 0.023 0.070 0.182 0.066
CS 0.386 0.052 0.013 0.077 0.030 0.112
CO 0.171 0.366 0.093 0.077 0.182 0.178
H 0.386 0.471 0.839 0.692 0.545 0.587
PF 0.014 0.105 0.031 0.077 0.061 0.058
∑ 1 1 1 1 1 1
After knowing the weighting of each criterion (also using DECISOFT 2.0), the TOPSIS or Ordered Preference method is deployed to determine the ranking of the five Restaurants. To begin with, an alternative criteria matrix is constructed that is applied to each of the five experts. Said matrix is completed using a Likert-type scale and the latter is detailed below:
Scale:
1: totally unsatisfactory
2: unsatisfactory
3: neither unsatisfactory nor satisfactory
4: satisfactory
5: totally satisfactory
After the five experts have responded to the matrix using the aforementioned measurement scale, a single matrix obtained with the use of the central tendency measure: median is made. That unique matrix is the following:
The non-existence of significant differences between the criteria or opinion of the experts is verified using the non-parametric Kruskal-Wallis test, assuming the assumptions:
- The data do not follow a normal distribution (free distribution) The data is processed with an ordinal scale (Likert)
The results of the test, obtained by processing the data using SPSS 11.5 software, are as follows:
Kruskal-Wallis Test
Sample Size Average Rank
--------------------
Col_1 5 10.2
Col_2 5 10.2
Col_3 5 12.4
Col_4 5 13.9
Col_5 5 18.3
--------------------
Test statistic = 5.02825 P-Value = 0.284411
The StatAdvisor
-----
The Kruskal-Wallis test tests the null hypothesis that the medians within each of the 5 columns is the same. The data from all the columns is first combined and ranked from smallest to largest. The average rank is then computed for the data in each column. Since the P-value is greater than or equal to 0.05, there is not a statistically significant difference amongst the medians at the 95.0% confidence level.
The Kruskal-Wallis test tests the null hypothesis that states that the means between the values of the five columns (which represent the criteria or attributes) are the same (H 0: Me 1 = Me 2 = Me 3 = Me 4 = Me 5 = Me). If the probability value is greater than or equal to 0.05 then there are no significant differences between the means. In the research, the probability value reached is 0.284411 (P> 0.05), hence it can be stated that there are no significant differences between the judgments made by the experts.
P CS CO H PF
R 1 0.331 0.331 0.419 0.371 0.404
R 2 0.442 0.442 0.314 0.464 0.505
R 3 0.442 0.442 0.419 0.464 0.404
R 4 0.552 0.552 0.524 0.464 0.505
R 5 0.442 0.442 0.524 0.464 0.404
Weighted normalized decision matrix using the weights calculated for each attribute using the Hierarchical Analytical Process method:
CO (0.178)
P (0.066) CS (0.112) H (0.587) PF (0.058)
R 1 0.022 0.037 0.075 0.237 0.022
R 2 0.029 0.050 0.056 0.296 0.027
R 3 0.029 0.050 0.075 0.237 0.027
R 4 0.036 0.062 0.093 0.296 0.027
R 5 0.029 0.050 0.093 0.237 0.027
Artificial alternatives:
*
H | 0.036 0.062 0.093 0.296 0.027 |
H - | 0.022 0.037 0.056 0.237 0.022 |
Separation of the ideal solution:
R 3 0.007 0.012 0.018 0.059 0.000 0.096
R 4 0.000 0.000 0.000 0.000 0.000 0.000
R 5 0.007 0.012 0.000 0.059 0.000 0.078
Separation of the negative ideal solution:
P CS CO H PF ∑
R 1 0.000 0.000 0.019 0.000 0.000 0.019
R 2 0.007 0.013 0.000 0.059 0.005 0.084
R 3 0.007 0.013 0.019 0.000 0.005 0.044
R 4 0.014 0.025 0.037 0.059 0.005 0.140
R 5 0.007 0.013 0.037 0.000 0.005 0.062
Relative proximity to the ideal solution:
R 1 0.1357
R 2 0.6000
R 3 0.3143
R 4 1.0000
R 5 0.4429
Descending order of preference:
R 4 1.0000
R 2 0.6000
R 5 0.4429
R 3 0.3143
R 1 0.1357
The restaurant with the best and most advantageous position in the ranking is R 4.
It was found during the development of the research that:
- The R 1 restaurant must improve in terms of the professionalism of its employees, which directly affects the service quality indicator, which also has deficiencies. To do this, the implementation of a training and education system in the hotel is proposed aimed at those employees who work in that restaurant, and even those who are about to work there. It is convenient for the Training department to prioritize the entry to the courses of personnel who in performance evaluations have had difficulties with the professionalism indicator. It is claimed that the quality of the offer is adequate in general as well as hygiene and physical presence of the restaurant dependientes.El R 2It presents notable deficiencies regarding the quality of the offer since the menu has not been focused on the eating habits of the customers who visit it, nor has it been correctly understood what the diners expect from the offer. For this, it is necessary to completely redesign the menu to provide during the service, the typical Creole and international food that customers demand and desire. Highlights that the restaurant has proper hygiene since it was built a bigger place in your kitchen area enabling a better scrubbing and cleaning the place during cooking of the restaurant alimentos.El R 3it has its greatest deficiency in the hygiene indicator since it has a very small and old kitchen that must be rebuilt as soon as possible so it does not have all the physical conditions created. Currently the hotel is involved in an investment project that alludes to the kitchen of this restaurant where the remodeling plans show that a better equipped area will be obtained that allows the restaurant to function and develop correctly. Despite the above deficiency, it is highlighted that the physical presence of its employees is quite good Restaurant R 4(as demonstrated by its first place in the ranking) it performs successfully in the installation since the professionalism of its workers, the quality of the service offered and the quality of the offer provided; They are good in general as well as the hygiene of the place and the food. The physical presence of the shop assistants is also adequate. Restaurant R 5It stands out for the correct preparation of the dishes it offers and its diversity as well as the physical presence of the employees who work there. It is emphasized that the quality of the service must improve in the same way that hygiene must be achieved from a drastic change in the place where this restaurant is located. It is stated that it is in a very small area for its capacity and that therefore the spaces for the kitchen, the waste area, etc. are reduced. so hygiene has been affected for some time. It is recommended (and the hotel had already conceived it) to open that restaurant in a larger place in accordance with the spatial standards for hotel catering places and also to offer,employee training courses on how to achieve hygiene, which is so important in the gastronomy area.