Application of Principal Component Analysis (PCA) to Determine the Catalysts and Hurdles for Efficacious Local Area Planning

Download Full-Text PDF Cite this Publication

Text Only Version

Application of Principal Component Analysis (PCA) to Determine the Catalysts and Hurdles for Efficacious Local Area Planning

Anika Kapoor

Research Scholar,

Amity School of Architecture and Planning, Amity University, Noida, Uttar Pradesh

Ekta Singh

Professor and Head,

Amity School Design, Amity University, Noida, Uttar Pradesh

Abstract While variables of participatory approach to urban planning been studied since long, many planners and researchers have often deserted the preliminary study of the variables acting as catalysts and hurdles for the implementation of local area plans. This paper adopted the principal component analysis (PCA) to assess variables by finding out the level of redundancy among them from the correlation matrix and grouping indicators with higher similarities into the same factors. We then estimated the unrotated and rotated factor loadings with PCA method. Results from the PCA were acceptable since we obtained extreme loadings after the varimax rotation. The results identified factors in terms of important hurdles and catalysts responsible for successful local area planning.

Keywords Principal Component Analysis, Local Area Planning, Participatory Approach

  1. INTRODUCTION

    Irrespective of the discomforts, constraints, difficulty, and cost, particularly in intricate, heterogeneous and advanced 21st- century cities, participatory local level planning as an approach, keep on drawing our attention as a resolution. The major intent and relevance of public participation at local level is identification of needs and concerns at local level to facilitate the institutions that are important for effective decision-making and urban development (Nabatchi, 2012). There are various models of planning at local level practiced across the world on a time line (Kathlene, 1991). Besides, there are numerous studies that focus on different forms of participatory plans, factors stimulating and inhibiting stakeholders initiatives, impact of stakeholders participation on final decisions, thereby yielding plentiful indicators that may have direct or indirect impact on the formal decision making administrative machinery (Lane, 2005; Lowndes et al., 2001). This often raises a question on the relevance of these indicators in terms of real time local level planning for urban management. There is no research study on the determination of critical catalysts and hurdles in successful implementation of local area planning mechanisms. The purpose of this study is to apply principal component analysis (PCA) technique for identifying keys factors (Bryant & Yarnold, 1995) out of many that are identified through literature review, that are expected to have impact on the efficacious local area planning in the ward context. The study is based on perception of local people of one ward of Delhi. Hauz Khas, ward no. 164, is selected randomly for the purpose of conducting a pilot study.

    To facilitate the study, PCA technique is used since it is considered to be one of the statistical multivariate methods based on eigenvector decomposition, which consolidates the variables in which there is interrelated smaller number of principal components (Jolliffe, 2002). These principal components are visualized, while holding the variation exhibited in the data set, as far as possible (Sratthaphut, et al., 2017). The subsequent section elaborates the methodology of the study.

  2. METHODOLOGY

    This study is based on the exploratory PCA of self- reported data. PCA was carried out to determine the strength of the correlation between variables pertaining to hurdles and catalysts for local area planning. As reported earlier, ward Hauz Khas, ward no. 164, is selected for the pilot study. Population of the selected ward is approximately 38000 (calculated based on electoral data, 2017, Election Commission of India). The sample size is calculated to be 380, with the confidence level of 95%. It is intended to constitute an unbiased sample of local people. The term Local people, mentioned in the paper, refer to residents, shopkeepers (formal and informal), visitors and members of residents welfare associations. The composition of the sample is illustrated in the Figure 1.

    Figure 1: Composition of sample respondents in ward Hauz

    Khas.

    The ward houses one unauthorised colony, part of two urban villages and eight planned colonies, which have approved layout plans. The authors have attempted to make representation of samples from each of the typology of the

    settlements. Besides, Figure 2 illustrate the characteristics of the sample in terms of gender, age structure and education level.

    Figure 2: Characteristics of Samples in Ward Hauz Khas

    Since this is a perception based study; a questionnaire based study was conducted to evaluate the perception, attitude and feedback of the respondents. The questionnaire was constructed based on identified indicators through literature review (Fraenkel & Wallen, 2006) that affect local area planning mechanisms (Alan Walks, 2011, Amado M.P, Amado et al. 2009, Castelloe, P. et al. 2002, Chado, J., & Johar, F. B. 2016, Chado, J. et al. 2016, Frediani, A. A. 2015, Garcia- Zamor, J. C. 2012 Healey, P. 1992, Kathlene, L., & Martin, J.

    1. 1991, Marzuki, A. 2015, Poplin, A. 2012, Stewart, T. R. et al. 1984 and Wilson K, et al. 2015) as shown in Table 1. Subsequently, the questionnaire was divided in 3 sections, namely, Section 1- Role of Government, Section 2- Willingness of stakeholders, Section 3- Role of technology. All the sections were further divided under the two categories Hurdles and Catalyst. A five-point adjectival ordinal scale was used to measure the responses (Strongly Disagree = 1, disagree

      = 2, Neutral=3, Agree =4, Strongly Agree = 5).

      After created the survey instrument, the validity and reliability was tested. Validity and reliability are interwoven as until a measure is reliable, it cannot be said to be valid. Cronbachs alpha is used to assess the reliability, or internal consistency. Cronbachs alpha is computed by correlating the score for each scale item with the total score for each observation (usually individual survey respondents or test takers), and then comparing that to the variance for all individual item scores.

  3. RESULTS AND DISCUSSION

    A list of 22 indicators is shown in Table 1. These were divided in three categories. Section 1 included behavioural indicators, section 2 encompassed of policy related indicators and section 3 comprised of technological indicators. The indicators under each section are further divided as hurdles and catalysts, as identified from Literature. In the table 1, Section is denoted by S, Hurdles are denoted by H and Catalysts are denoted by C.

    Table 1: List of Identified indicators of Local Area

    Planning.

    6

    S. No

    Code

    Indicator

    1

    S1_H_1

    MCD initiatives for development at local level are not satisfactory.

    2

    S1_H_2

    Local People are not given opportunity to share their viewpoint with respect to development of your locality in any case.

    3

    S1_H_3

    MCD is biased towards rich and influential.

    4

    S1_C_1

    Core function of MCD is Planning and monitoring of local areas

    5

    S1_C_2

    The prime focus of MCD should be Service delivery

    S1_C_3

    MCD should act as a link between people and state.

    7

    S2_C_1

    Local residents wish to participate in decision making at local level.

    8

    S2_H_1

    Illiteracy is the reason for reluctance for participation in decision making.

    9

    S2_H_2

    People do not participate in plan making due to lack of awareness

    10

    S2_H_3

    People think it to be waste of time and energy

    11

    S2_H_4

    People believe that their inputs will not be incorporated in the plan.

    12

    S2_H_5

    People have lost trust on local authorities

    13

    S2_C_2

    People are bothered about the development of their locality.

    14

    S2_C_3

    Local people are ready to pay minimum amount for development of local areas, beyond taxes.

    15

    S3_C_1

    Consumerist method is the most adopted means of public consultation (complaints and suggestions).

    16

    S3_C_2

    People are more comfortable with conventional methods of participation.

    17

    S3_C_3

    Technology is more widely accessible to citizens through internet, smartphone applications, and social media.

    18

    S3_H_1

    IT illiteracy of the government officials is the reason for reluctance in adopting technology.

    19

    S3_H_2

    IT illiteracy of the masses is the reason for reluctance in adopting technology.

    20

    S3_C_4

    Cost of adopting technology for public participation is not high.

    21

    S3_C_5

    Access to technology facilitates adoption of technology for participation.

    22

    S3_H_3

    Technology is not adopted for decision making due to Behavioural issues

    Next, Cronbachs alpha is calculated to measure the reliability of the variables in the questionnaire. It is recommend to have a minimum coefficient between 0.65 and 0.8 (or higher in many cases); coefficients that are less than 0.5 are usually unacceptable (Cortina J., 1993). The literature suggests that in case based heterogeneous questions, alpha should be calculated for each case, for efficacious results (Cohen, 2010). Therefore, Cronbachs alpha value is calculated for each section of the questionnaire to check the consistency. The results are as below in Table 2:-

    Table 2: Results of Cronbachs alpha

    Abbreviation

    Sections of Questionnaire

    Cronbachs Alpha

    S1_H

    Section 1: Hurdles

    0.751

    S1_C

    Section 1: Catalysts

    0.814

    S2_H

    Section 2: Hurdles

    0.691

    S2_C

    Section 2: Catalysts

    0.706

    S3_H

    Section 3: Hurdles

    0.717

    S3_C

    Section 3: Catalysts

    0.703

    Cronbachs alpha values showed fair to good internal consistency in all the three sections. This means that the questionnaire is reliable and good to proceed for PCA.

    As a first stage to PCA, the sampling adequacy can be assessed by examining The Kaiser-Meyer-Olkin (KMO) (Kaiser 1970). KMO varies between 0 and 1. The values closer to 1 are better and the value of 0.6 is the suggested minimum. This test provides the minimum standard to proceed for Factor Analysis. Normally, 0 < KMO < 1. If, KMO > 0.5, the sample is considered to be adequate (Hair, Anderson et al. 1995a). Section 1 and 2 show, KMO > 0.5, which indicates that the sample is adequate and we may proceed with the Factor Analysis. Bartletts Test of Sphericity was also applied. Bartletts test of Sphericity provides a chi-square output that must be significant (Bartlett 1950). Taking a 95% level of Significance, the results are 0.05. The p-value (Sig.) of .000 < 0.05, therefore the Factor Analysis is valid. Since the results of section 3 shows that the matrix is not positive definite, the KMO and Bartlett's Test table was not showing in Statistical Package for the Social Sciences (SPSS) software. The PCA results are nevertheless valid (Ulanowicz, 1986). Therefore, Factor Analysis is considered as an appropriate technique for further analysis of the data. Details of 1 section are elaborated and similarly results for other two sections are obtained.

    Table 3 shows Eigen values of section 1. This is one of most fundamental and most useful concepts in linear algebra. To explain further, if n × n matrix is called A, then A scalar is said to be a eigenvalue of A. The vector x is called an eigenvector corresponding to . Eigen Values >= 1 were only selected in the SPSS.

    Table 3: Eigen values of Section 1

    In the present section, only the 2 factors were extracted by combining the relevant variables. The Eigen values are the variances of the factors. The total column contains the Eigenvalue. 2 factors are extracted by combining the relevant variables. The first factor always accounts for the most variance and hence has the highest Eigen values. The next factor will account for as much of the left over variance. Each factor is constituted of all those variables that have factor loadings greater than 0.5. The percentage of variance represents the percent of total variance accounted by each factor and the cumulative percentage gives the cumulative percentage of variance account by the present. In this section, the first 2 factors explain 76.309% of variance.

    The scree plot graphs the Eigenvalue against the each factor. The graph shows that after factor 2 there is a sharp

    change in the curvature of the scree plot. This shows that after factor 2 the total variance accounts for smaller and smaller amounts. (Reference Figure 3)

    Figure 3: Scree Plot- Section 1

    Identification of the Core Factors

    The Rotated Factor Matrix characterises the rotated factor loadings demonstrating correlations between the variables and the factors as shown in Table 4. The factor column denotes the rotated factors that have been mined out of the total factor. These are the core factors, which have been used as the final factor after data reduction.

    Table 4: Rotated Component Matrix- Section 1

    The factors are grouped and each group of factors is named which is the representation of the grouped factor. Two factors explain the variance in performance of the variables in section

      1. Similarly, factors in section 2 and 3 were also identified. Results of Principal Components Analysis (PCA) of Section1, 2 and 3 are showed below in Table 5.

        Table 5: Identified Significant Factors in Section1, 2 and 3-

        Catalysts and Hurdles

        S.no

        Factors

        Description of the factor

        POLICY FRAMEWORK INDICATORS

        1

        S1_H1

        People at local level are not satisfied with the MCD services and feel deprived of opportunity to participate in the decision making for development of their locality.

        2

        S1_C1

        MCD should be dedicated to deliver and monitor services, thereby realizing the needs/problems of the people at local level and extending them to state for

        prompt resolutions.

        BEHAVIOURL INDICATORS

        3

        S2_C1

        Literate local residents are willing to participate in the decision-making for development of the locality.

        4

        S2_H1

        Lack of awareness is the reason for non- participation of people at local level.

        5

        S2_H2

        People perceive public participation to be waste of time and energy, since their inputs are not considered while planning for local areas.

        6

        S2_H3

        People are concerned about development of their locality but have lost trust on ULBs, because of their

        past experiences.

        TECHNOLOGY INDICATORS

        7

        S3_C1

        People are comfortable with conventional methods of participation but are ready to adopt the technology through various means.

        8

        S3_H1

        IT Illiteracy and reluctance of the officials to adopt the new means of participation strategy are the reason for not adopting technology.

        9

        S3_C2

        People are ready to adopt technology as it is cheap and

        accessible.

  4. CONCLUSION

This paper aimed to investigate into various attributes related to local area planning. PCA technique is the powerful choice for the determination of most significant factors which acts are hurdles and catalyst for implementation of Local Area Plans. 22 indicators as shown in Table 1 were identified from literature. 11 indicators were pertaining to hurdles and 11 refer to the catalysts to local area planning mechanism. Subsequently, the number reduced drastically to 9, where 5 indicators point to hurdles and 4 point toward catalysts for successful implementation for local area planning mechanisms in the ward context. In addition, this conclusion is also useful for policy makers and practitioners in highlighting these attributes to refine adaptable local area planning framework in order to achieve sustainable solution to urban development at local level. This work can be extended to investigate whether these indicators are correlated with the characteristics of different wards in Delhi and its adjoining cities.

REFERENCES

      1. Alan Walks, R. Planning sustainable cities: global report on human settlements 2009 by United Nations Human Settlements Programme. Earthscan/UNHabitat, London, 2009. No. of pages: xxx+ 306. ISBN 978 1 84407 899 8. Population, Space and Place, 17(3), 2011, 290- 292.

      2. Amado M.P, Santos C.V., Moura E.B and Silva V.G. Public Participation in Sustainable Urban Planning" International Journal of Civil, Environmental, Structural, Construction and Architectural Engineering Vol:3, No:5, 2009, 2009, Pg. No. 597-603

      3. Bartlett, M. S. (1950). "Tests of significance in factor analysis. British Journal of Psychology 3(2): 77-85.

      4. Bryant, F. B., & Yarnold, P. R. Principal-components analysis and exploratory and confirmatory factor analysis. In L. G. Grimm & P. R. Yarnold (Eds.), Reading and understanding multivariate statistics (1995, pp. 99-136). Washington, DC, US: American Psychological Association.

      5. Castelloe, P., Watson, T., & White, C. Participatory change: An integrative approach to community practice. Journal of Community Practice, 10(4), 2002, pp 7-31.

      6. Chado, J., & Johar, F. B. Public Participation Efficiency in Traditional Cities of Developing Countries: A Perspective of Urban Development in Bida, Nigeria. Procedia-Social and Behavioral Sciences, 219, 2016, pp. 185-192.

      7. Chado, J., Johar, F. B., & Zayyanu, M. Challenges Impeding Public Participation for the Development of Urban Communities in Nigeria. Indian Journal of Science and Technology, 2016, pp. 9(46).

      8. Cohen R, Swerdlik M. Psychological testing and assessment. Boston: McGraw-Hill Higher Education; 2010.

      9. Cortina J. What is coefficient alpha: an examination of theory and applications? Journal of applied psychology. 1993; 78:98-104.

      10. Fraenkel, J. R. & Wallen, N. E. How to design and evaluate research in education (6th ed.). Boston, MA: McGraw Hill, 2006.

      11. Frediani, A. A. Participatory Capabilities in Development Practice. The Bartlett Development Planning Unit, 2015, pp. 121.

      12. Garcia-Zamor, J. C. Public participation in urban development: The case of Leipzig, Germany. Journal of Public Administration and Policy Research, 4(4), 2012, pp. 75.

      13. Hair, J., R. Anderson, et al. (1995a). Multivariate data analysis. New Jersey, Prentice-Hall Inc.

      14. Healey, P. Planning through debate: the communicative turn in planning theory. Town planning review, 63(2), 1992, pp.143.

      15. Jolliffe IT. 2002. Pricipal Component Analysis 2nd Ed. New York, USA: Springer-Verlag.

      16. Kaiser, H. F. (1970). "A Second-Generation Little Jiffy." Psychometrika 35(4): 401-415.

      17. Kathlene, L., & Martin, J. A. Enhancing citizen participation: Panel designs, perspectives, and policy formation. Journal of Policy Analysis and Management, 10(1), 1991, pp. 46-63.

      18. Lane, M. B. Public participation in planning: an intellectual history. Australian Geographer, 36(3), 2005, pp. 283-299.

      19. Lowndes, V., Pratchett, L. and Stoker, G. "Trends in Public Participation: Part 1 – Citizens' Perspectives", Public Administration, 79(2), 2001, pp. 205-222

      20. Lowndes, V., Pratchett, L., & Stoker, G. Trends in public participation: part 2citizens' perspectives. Public administration, 79(2), 2001, pp.445-455.

      21. Marzuki, A. Challenges in the Public Participation and the Decision- Making Process. Sociologija i prostor, 53(1 (201)), 2015, pp. 21-39.

      22. Nabatchi, T. Putting the public back in public values research: Designing participation to identify and respond to values. Public Administration Review, 72(5), 2012, pp. 699-708.

      23. Poplin, A. Playful public participation in urban planning: A case study for online serious games. Computers, Environment and Urban Systems, 36(3), 2012, pp. 195-206.

      24. R. E. Ulanowicz: Growth and development: Ecosystems phenomenology. New York, SpringerVerlag, 1986, 203 pp.

      25. Sratthaphut L, Tanyasaensook K, Tunneekul P. Principal Component Analysis: A Tool for Identifying Web Document Characteristics Affecting Quality of Drug Information Websites. J App Pharm Sci, 2017; 7 (11): 001-006.

      26. Stewart, T. R., Dennis, R. L., & Ely, D. W. Citizen participation and judgment in policy analysis: A case study of urban air quality policy. Policy Sciences, 17(1), 1984, pp. 67-87.

      27. Wilson.K, Hannington.S, Stephen.M The Role of Community Participation in Planning Processes of Emerging Urban Centres. A study of Paidha Town in Northern Uganda. International Refereed Journal of Engineering and Science (IRJES), Volume 4, Issue 6 (June 2015), 2015, PP.61-71.

Leave a Reply

Your email address will not be published. Required fields are marked *