"certain ways, and is sometimes called behavioural component. Maloney et al.?s (1975) researchfirst advocated importance of determining such components.These authors built a scale with four subscales. Initially, the original scale was composed of 128items on an ecological attitude-knowledge scale, but they revised to provide a more practical andefficient instrument. During research author left 10 items in the scale that measured emotionalcomponent; 15 questions in the factual knowledge scale about environmental cognitive features;a 10-item verbal commitment scale that measured positive and negative environmental attitudeand finally, a fourth scale that assessed ecological behaviour with 10 items. At the beginning,three environmental attitude components – affective, cognitive and conative – are used to predictecological behaviour. However, we found other approaches that show one of those componentsas single indicators of environmental attitude (Arbuthnot, 1977; Dispoto, 1977; Schahn andHolzer, 1990). A “DEM ” (Direct effect model) was developed with these components whereaffective and verbal commitment was considered as independent variables and, conversely on theother hand, actual commitment or intentional component was considered as dependent variable.Verbal commitment has been used as a mediator to describe attitude towards behaviour.3.3 The proposed model and hypothesesUsing theoretical basis for organizing relevant conceptual model outlined before, a three- dimensional „attitude approach? was developed concerning attitudes towards environment andecological behaviour. This study seeks to define relationship between three components (Viningand Ebreo, 1990) (Fig. 1): affect commitment (Af), verbal commitment (VC) and actualcommitment (AC). The Theory of Reasoned Action argues that attitude influences behaviorthrough intention, although these variables affect and influence behaviour directly. Previous6 studies found this relationship become stronger even (Dunlap and Van Liere, 1978; Ajzen andFishbein, 1980; Chan, 1999, 2001; Kaiser et al., 1999a,b) while discussedwith respect toecological behaviour. In this analysis, AC was considered as behavioural component (Bhate andLawler, 1997).The literature demonstrates that relationship between values, which make people to show ahigher or a lower affection level towards something or someone control environmental attitudes(Schultz and Zelezny, 1999).Incidentally several studies have already demonstrated relationshipbetween values on this scale and environmental behaviour (Stern et al., 1995; Thogersen andGrunert-Beckmann, 1997)Ecological behavior- AffectActual commitmentcommitment(AC)H2 (Af)Environmental attitude-VerbalcommitmentH H(VC) 3 1 H3 Figure1-The proposed model of our research (DEM-Direct Effect model)- Partial mediation7 On the other hand, Maloney et al.?s (1975) opined that revised scale and its affect on differentsubscale is a predictable phenomena, Nowhere author can distinguish between those studies thatfound a declining relationship between individual?s environmental affect and ecologicalbehaviour because there is another mediator variable in-between, which is none other thanecological behaviour or attitude (Ajzen and Fishbein, 1980; Kotchen and Reiling, 2000). So,based on depicted relationships in Fig. 1 related to environmental affect and on literaturefindings, following hypotheses have been developed:Hypothesis1: People with a higher degree of affect about the environment will have strongerenvironmental attitude.Hypothesis 2: People with a higher degree of affect about the environment will have strongerecological behavior.Hypothesis 3: People with a major environmental attitude will demonstrate a strongerecological reaction.4.0MethodologyA pre-test questionnaire was administered to 50 citizens of the Kolkata city. This pre-test (Pilotstudy) was developed to discover any possible weakness in the questionnaire. It was revised and,th between 7th Jan to 7 April 2014, a final questionnaire was administered to a random sample of384(n ) individuals. From them, a total of 360 valid (93.75%) questionnaire was selected for0 final analysis. As the sample is drawn from a finite population so a specific assumption wasmade that p- the maximum variability inpopulation proportion is considered as 0.5 so the valueof q will be 0.5, The sample size (n) is calculated as follows2 2 n= (Z pq)/e .where z =1.96 and e (0.05) at 5% level of significance. Further sample sizecorrection is done by the following method.8 n (Nn)/ (N-n+1), where N is the assumed population size of Kolkata metro city (census report0= 2001). Non probabilistic sample process (convenient sampling) is used to collect primary datafrom market. Researcher delivered questionnaire to householders and collected once they arecompleted. The questionnaire consisted of three large blocks. The first portion includedquestions about recycling practices and the purchase of ecological products. The second blockattempted to measure variables such as personality, attitudes, values and individuals? lifestylesand the third asked about demographic and socio-economic factors. Table no1 Sample profile of the survey DemographicsFrequency-N=360PercentageFemale201 0.558333Male159 0.441666(Age) 15-24102 0.28333325-2996 0.26666640-5463 0.17555-6460 0.166667>6439 0.108333Without studies Primary school63 0.175Secondary school195 0.541666667University and above102 0.283333333OccupationStudent75 0.208333333Housewife51 0.141666667Entrepreneur/professional66 0.266666667Executive/independent worker36 0.19 Manager33 0.091666667Specialized worker12 0.033333333Retired27 0.075Unemployed24 0.066666667Others6 0.016666667Family Income Below10, 000 48 0.13333333310001-20,000 156 0.43333333320001-30,000 144 0.316666667above 30,00042 0.116666667 Each scale originally had 10 items and respondents had to indicate if the statement was true orfalse. This was transformed into a 7-point Likert scale that ranged from „disagree with all? (1) to„strongly agree? (7). After pre-test, three items were removed from ecological Af and VC scales,and one from the AC scale. A brief scheme about number of people who totally agreed or totallydisagreed with each statement for these three scales can be seen in Table 2.This table shows thatfor the environmental affect scale, a high percentage of people were concerned aboutenvironmental problems. The chronbach alpha value (mentioned below) on 22 items indicates(0.868>0.7) that almost 86% of the collected primary data are considered to be reliable foranalysis.Reliability StatisticsCronbach's Alpha N of Items.868 2210 "