AUTHOR’S DETAILS:
1. Name: Dr.L.Sudershan Reddy
Professor, Faculty of management, Jain University, Bangalore.
Mobile No : 9845472096, Email id:
2. Name: Kannamani R
Doctoral scholar, Faculty of management, Jain University, Bangalore.
Mobile No : 9986595820 , Email id:
Correspondence Address:
Kannamani Ramasamy,
2nd floor, #33, 3rd cross, Munireddy layout
Near Matha school, Garapalaya, Bangalore 560068
State: Karnataka, Country: India
Mob: 9986595820
A Pilot Study– An analysis and effective implementation of matrix organization in IT outsourcing industry
L. Sudershan Reddy[1], Kannamani R[2]*
Abstract:
Matrix organization is a contemporary management concept in IT outsourcing industry. Pilot studies are a critical component of a quality study design. In general, pilot study is considered as a miniature version of a complete research study. Conducting a pilot study does not guarantee success in the actual research, but it helps to improve the probability of success. In this research paper, we discuss few important things such as the need of the pilot study in IT outsourcing industry with respect to matrix organization, the process of testing the feasibility of the project proposal, requirement of sample, research tool, data analysis and the outcome of the study. Based on this pilot study we are going to identify the parameter which needs to be fine-tuned in order to carry out the research with large number of participants.
Keywords: Matrix organization, IT outsourcing, Infrastructure technology outsourcing, Respondents, Reliability, Validity, Service delivery, Hypothesis, Sampling, Population.
Abbreviations: ITO –Infrastructure technology outsourcing, MOS- Matrix organization system, IT- information technology,
Introduction:
Matrix organization is a concept where an employer managing an employee by using more than one manager or in other words we could say an employee reporting to more than one manager in order to obtain an organization delivery expectation which is part of organizations vision and mission. Any organization outcome depends on the reporting relationship with the management and employee. Hence strategy for employee management becomes very critical.
Earlier organizations were using conventional management concept where an employee reports to a supervisor, supervisor reports to a manager, manager reports to his manager (Say general manager), and general managers reports to managing director. In this method request goes from bottom level (from employee level) and command comes from top level (management level, say managing director).May be this concept was good when they had legacy production system, manual operations, partially skilled employees, less automation, with less number of employees. But when the changes are needed in all or any of the parameters it is important to have management concepts as per the need.
As per the review of literature, Matrix structure was employed in 1947 by General Chemicals in its engineering department and also in 1960s the matrix method called "unit management” was implemented by a large number of U.S. hospitals. But however the first organization to frame, design, develop and implement a formal matrix structure was NASA. NASA developed a matrix management system for their space program since it required to instantaneously emphasizing several different functions and projects, none of which could be stressed at the expenditure of another. They have identified that the traditional management system were not suitable because it is too bureaucratic, hierarchical, slow-moving, and inflexible.
Because of the effectiveness and benefits to the organizations and business, matrix management gained broad acceptance in the corporate world during 1970sand recognized largely during 1980s as a result of economic changes in the United States, which comprised slackening national market progress and increasing overseas competition. Those fluctuations forced many establishments to pursue the benefits offered by the matrix organization model.
Limitations of the Study:
· This pilot study is restricted to IT outsourcing industry (Infrastructure technology outsourcing)
· This pilot study is concerned upon randomly selected 10+ IT outsourcing companies across the globe.
· This study is restricted to companies which are either in to CMM level 5 or fortune
· The study examines the opinion of employees, project managers and functional managers from various angles.
· This study aims to understand the MOS as to how it is being implemented and used by the employer and employees in order to find the demerits of it.
Uniqueness of the Study:
· The existing research on MOS is more on Non-IT industry and we see very little research from IT perspective. The uniqueness of this research is that the research has been conducted in top ranking organizations in the IT outsourcing industry.
Research approach and methodology:
Background:
Matrix organization systems are important and unavoidable concepts in IT outsourcing industry. We see very few researches happening in IT industry, also when it comes to ITO industry even it is very less and also the context of the researches are not up to the mark to overcome the difficulties which are being faced by the employees and employer. Based on the personal experience with various ITO MNCs, the way of implementing the MOS in real time is different from what was defined in white & black.
This research aims to find out the challenges that are exist in IT outsourcing industry with respect to matrix organization and fine out how we can further fine-tune and give the remedy for the challenges.
Need of pilot study in this research
· Verify that the directives given to respondents are understandable.
· Check the reliability and validity of the results.
· Measure whether the level of intervention is appropriate.
· Developing and testing suitability of research instruments
· Judging the practicality of a the study/survey
· Designing a research protocol
· Assessing whether the research protocol is realistic and workable
· Forming whether the sampling frame and technique are effective
· Assessing the likely success of projected recruitment approaches
· Identifying logistical challenges which might occur using proposed methods
· Estimating variability in outcomes to help determine the sample size
· Gathering preliminary data
· Determining what resources are needed for a research study
· Calculating the proposed data analysis techniques to uncover probable complications
· Developing/Modifying a research question and research plan
Research objectives:
· This pilot study aims to identify the effectiveness of matrix organization system and the various factors associated with it.
· To find out the challenges those are being faced by an employee and an organization in real time due to matrix organization system.
· To create a developed or fine-tuned process in order to overcome the challenges, this will give benefits to the employee, IT customer and an organization.
Research Design:
· Type of research - Quantitative method
· Research methods used - Questionnaire based Survey
· People interviewed -An ITO employee includes engineer level, manager Level, senior manager level, project managers.
· Numbers of interview -44 (out of 55 target population )
· Research locations - Global
· Questionnaire -Based on 5 hypothesis, it was divided in to 23 questions
· Method of data collection - Structured questionnaire sent via email.
Sampling technique used and why:
We have chosen judgment sampling which comes under non probability sampling technique. The reason for choosing this method is that we are in to pilot phase, uncertainty of questions and responses, targeting specific audience and also for ease of operations. Based on the outcome we may try to pick some other sampling methods if needed.
Survey instrument:
Email is our pilot study research instrument .In every email communication with the respondents we have given an introduction about the research scholar and need of doing this research. Questionnaire has been prepared in the google forms and the form has an introduction about the matrix management and the purpose of asking the response from respondents which will help them to understand the subject and correlate with their experience when they are answering the questions so that we will get an answer with more relevance. All the responses will be stored in the google spread sheet where we can analyses the data later.
Organization and participants:
We had 44 participants from various organizations. They are all holding different positions such as engineer, project manager, Functional manager and senior management professionals. Respondents are from various continents such as Asia, Europe, North America and South America. Participants are from various countries such as India, Philippines, Malaysia, USA, England, Mexico, Australia and Guatemala. Importantly we have chosen participants from various top IT outsourcing companies such as IBM, HP, ATOS, CTS, EMC, Microsoft, TCS, Tech Mahindra, Vodafone, Wells Fargo, and Xerox.
Hypothesis and questionnaire:
We are intended to test the hypothesis. In order to get the data from the respondents we have divided our 5 hypothesis in to multiple questions. In total we have asked 27 questions, 22 belong to hypothesis which will help us to collect the data in order to analyse and draw the inferences. 5 questions are asked in order to understand the respondents details such as position, company name, country name, continent name and number of years’ experience in IT industry which helps to understand the participants statistics and it helped to prove that the respondents are not from same organization, not from same country and not in the same experience group. This also helps us to justify the quality of the representative sample.
Hypothesis 1: Employee level issues (Performance rating, Job security, Job Comfort)
Hypothesis 2: Continuous improvement plan
Hypothesis 3: Conflict issues
Hypothesis 4: Responsibility and authority issues
Hypothesis 5: Global issues (diversity, ethnic variance, language and global position/pay)
Data collection and response rate:
We had distributed questionnaires to 55 respondents and worked with them for 1 month in order to ensure that they are responding to the questions. Out of 55 target population, we have got response from 44 people which is around 80% of the total crowd.
Tools and techniques for data analysis:
We have used IBM SPSS statistics 22 tool in order to analyze the descriptive statistics such as “Frequencies and descriptive, Dimension reduction- Factor analysis for knowing the validity of the questionnaires and , cronbach's alpha-reliability analytics for analyzing the reliability or consistency of the questionnaire and one sample T test to test the hypothesis”.
Descriptive statistics:
Hypothesis testing:
Non parametric hypothesis testing (One-sample chi-square test) used here and based on the decision table we can conclude that we may have to reject null hypothesis (Asymptotic significance vary from 0.000 to 0.11)
Reliability test:
To identify the reliability of the pilot study, we have to run the cronbach’s Alpha test. The reliability score indicates that the collected data is unbiased. The reliability of the study is good as the cronbach's alpha score is 0.877.
In general, (As per Wikipedia, as on 4 January 2016, at 15:21) Cronbach’s alpha increase as the intercorrelations among test items increase, and is thus known as an internal consistency estimate of reliability of test scores. Because intercorrelations among test items are maximized when all items measure the same construct, Cronbach's alpha is widely believed to indirectly indicate the degree to which a set of items measures a single unidimensional latent construct. It is easy to show, however, that tests with the same test length and variance, but different underlying factorial structures can result in the same values of Cronbach's alpha.
Suppose that we measure a quantity which is a sum ofcomponents (K-itemsortestlets):. Cronbach'sis defined as
Whereis thevarianceof the observed total test scores, andthe variance of componentifor the current sample of persons
The theoretical value of alpha varies from zero to 1, since it is the ratio of two variances. However, depending on the estimation procedure used, estimates of alpha can take on any value less than or equal to 1, including negative values, although only positive values make sense. Higher values of alpha are more desirable.
Method 1: Alpha
Method 2: Guttman split half coefficient
Validity test:
In order to identify the validity of the questionnaire we have run the factor analysis by using “Principle component, axis factoring, Image factoring methods” and result as below:
Missing value analysis:
It is very critical to understand the missing values which may impact the outcome of the study. In order to get the expected outcome we need to ensure the respondents are answering all the questions. Below is the missing value pattern of out pilot study.
Remarkable Correspondence analysis:
In order to identify and understand the connections which shows the similarity and controversial outcome of the questionnaire we ran correspondence analysis and output given below and discussed in details in “discussion” area.
Discussion:
1. Hypothesis can be modified as the effectiveness comes from various factors.
2. Performance rating question can be modified as “Are you fine to rate your performance by someone, other than who manage your day to day work?”
3. Identified 7 missing values and we need to instruct the respondents to answer all question so that the validation part will be good.
4. The reliability and validity test results are within the expected range but however the factor analysis indicated the value as 0.56 which can be improved may be by fine tuning the questions slightly as per the research objectives.
5. Employee position needs to be classified with more items
6. Correspondence analysis indicates that there are few components which are contrasting the result of another component which is in the same area.
7. When we asked respondents that “Are you comfortable with MOS?”,30 people responded as “yes” but when we asked about the need of centralized performance management system,34 people said it is needed. Though they are comfortable with MOS, but based on their experience they want to have CPMS.
8. When we asked respondents that “Are you comfortable with MOS?”,30 people responded as “yes” but when we asked about the need of skip level approach,31 people said it is needed. Though they are comfortable with MOS, but based on their experience they wanted to have skip level approach system. Few of the companies are having it already, but however some of them are not using effectively and some of the organizations yet to implement.
9. 22 people only answered that they are happy with the performance rating which is 50% of the sample
10. 36 people mentioned that they need continuous improvement plan (CIP) in their organization.