Spss tpb

1

Note that this assignment is under 1,600 words so there is room for expansion.

Introduction

An investigation of the behaviour of smokers and their attitude toward quitting can

deliver valuable information to create effective measurements to work towards a

lower smoking prevalence in the future. Within this study the theory of planned

behaviour (TPB) is applied, with the purpose to investigate the influence of the TPB

antecedents on intention (INT) of smokers to quit smoking. The antecedents are

defined as attitude (ATT), subjective norm (SN) and perceived behavioural control

(PBC). A key focus within this research is to discover similarities and differences

across the two different European countries: Greece and Spain.

The paper will first give a theoretical overview about the theory of planned behaviour.

Secondly the sample profile will be briefly discussed. A descriptive analysis will

discuss differences between the mean scores of the antecedents across the two

countries. Based on the findings a regression analysis will demonstrate the

predictability of the intention to quit smoking by the TPB variables and how this

differs between the nations. Finally a conclusion will give ideas for the usefulness of

the acquired data and discuss research limitations.

Theory of planned behaviour

The TPB is an expectancy value model which states that human behaviour is a

consequence of one’s behavioural intention, which is in turn explained by one’s

attitude, subjective norm and perceived behavioural control regarding the behaviour.

The TPB is an extension of the theory of reason action (TRA; Fishbein & Ajzen,

1975) which does not include perceived behavioural control and thus not designed to

explain behaviours out with an individual’s volitional control. In the TPB and TRA

models, behavioural intention is the cognitive representation of the individual’s

motivation to enact the behaviour in question. Attitude toward the behaviour is the

individual’s positive or negative evaluation of performing the behaviour. Subjective

norm is the individual’s perceived social/peer pressure in undertaking the behaviour.

Lastly, perceived behavioural control represents the individual's perception of the ease

or difficulty in enacting the behaviour. The TPB has been found to explain intention

Comment [E1]: Executive summary needed. This is part of the assignment

requirements. Should also include title and

contents page.

An executive summary should provide a

summary of all the research process. i.e. what the objectives of the research are, how

the research was conducted, the key results

and recommendations.

Comment [E2]: Could be more specific as to why this is important, including some

literature.

I would expect research objectives to be clearly stated, with associated research

questions to be answered

Comment [E3]: What about any specific studies in the content of smoking cessation?

These would be very relevant to refer to.

Reference to key journal articles pertinent

to the research topic is needed.

For your report, there is no need to dwell on

theory, but you need to give some brief background on the key variables such as

perceived value in the appendix.

2

and behaviour in a wide variety of contexts and is considered to be the most used

model in understanding health behaviour (Astrom & Rise, 2001).

Analysis and Results

Sample profile

The study was conducted through an online questionnaire in Greece and Spain.

Overall 248 responses were received with a higher proportion (60% or 150) of

responses from Greece and the remaining 98 responses from Spain. A number of

questions were used to collect information about the sample, these were gender, age

(in years), household income (in €), and social position. All respondents identified

themselves to be smokers. Figure 1a and 1b show the proportion of respondents who

are male and female for each country. There are proportionally more males for the

Greek sample than in the Spanish ample. A Chi-square test showed a significant

difference in proportion of males/females across country (p < .05). Figure 2a and 2b

gives the social position of respondents for each country. The bar charts show that for

both countries around half of the respondents are in full-time employment. For Spain

there are a higher proportion of respondents across each of the remaining categories

than for Greece. For the Greek sample most respondents who are not employed full-

time are in the categories of student or other. The other category may represent

homemakers but this is not clear from the data provided. In terms of age and income

across the countries, Table 1 provided descriptive statistics. The mean values

highlights that the Spanish sample are overall older than the Greek sample. An

independent samples t test confirmed that the mean age of respondents is different (p

< .05) across the two countries. Difference in mean household income was also

statistically significant (p < .05) indicating that the levels of household income differ

across the Spanish and Greek respondents. The Greek respondents have on average a

higher household income than the Spanish respondents.

Comment [E4]: Could also include a descriptive analysis of the variables, e.g.

was intention to quit high/low?

Comment [E5]: Avoid mixing present and past tenses in your reporting. You can

use either present or past tense, but keep

consistently to one choice.

Comment [E6]: A vague term. Try to avoid ambiguous terms. State clearly what

you mean.

Comment [l7]: There is a higher proportion

Comment [l8]: These two statements contradict each other. The statement for

Spain is wrong and does not reflect the

views shown in Figures 2a and 2b.

3

Figure 1a. Pie Chart showing Gender for

Greece

Figure 1b. Pie Chart showing Gender for

Spain

Figure 2a. Bar Chart showing Social Position for Greece Comment [E9]: Should use a clustered bar chart instead of two separate charts.

4

Figure 2b. Bar Chart showing Social Position for Spain

Table 1. Descriptive statistics for age and income across Country

T-tests across Country

It is important to determine if the two countries are similar in regard to the mean level

of the TPB variables. Before discussing the t-test it is important to establish if the

respondents from each country have a similar variability in their responses to the TPB

questionnaire items. Results from the Levene’s test of equality of variances shows

that for the intention items and the overall average of intention there are significant

Comment [E10]: It is better to include the SPSS output as an appendix and create

your own tables using word. Your own table should contain only the centrally

important statistics (and also decide on

suitable number of decimal places). I suggest 2 dp’s. A table such as Table 1 of

Hassan and Shiu (2006) but giving the

number of items, the range, the mean and standard deviation for each construct would

be useful. It is important that the reader

understands what, for example, a mean of 2.5 represents - does this value perhaps

indicating a very positive attitude or a

moderately positive attitude? It also depends of the measurement i.e. how many

scale points are used and what the end

points of the scale are so this is important information.

Comment [E11]: Again see Hassan and Shiu (2006) for a further example of displaying and writing about t-test results.

Comment [E12]: This is vague! Why is it important and to whom is it important?

5

differences (p < .05) in the variability across the countries with respondents from

Spain being less homogenous in their responses than the respondents from Greece.

The levene’s test results for subjective norm show that there is no difference (p > .05)

in the variability of responses from respondents across the two countries.

In order to establish if there are mean differences an independent t-test was

conducted. The results of this analysis are presented in Tables 2 and 3. The results

show overall that there are many mean differences across the countries in terms of

intention and attitude. But that SN and PBC do not differ significantly across the two

countries. Importantly plan and intend significantly differ across the countries with

respondents from Spain having a higher intention to quit smoking. However

respondents in both countries both exhibit a similar mean regarding wanting to quit.

There are no differences across the countries in terms of the cognitive attitudinal

items, however there are differences across the countries in terms of the affective

attitude items. Respondents in both countries consider smoking to be negative and

bad. However smokers from Greece perceive smoking to be more pleasant and

enjoyable than respondents from Spain. This might represent a barrier to engaging in

quitting. In terms of SN respondents from both countries express some normative

pressure to engage in quitting. However the overall mean value is low. Respondents

from both countries perceive that quitting is within their control and therefore have

high mean values on the measures of self-efficacy.

Table 2. Means for each TPB variable across Country

Comment [E13]: Within the discussion of the t-test results, it is important to demonstrate an understanding of the

Levene’s test. The red text adds a

discussion of Levene’s test for intention and subjective norm only.

Comment [E14]: Would also be good to add in brackets the mean values so that readers can see some detail without having

to look at the table. For example, in terms

of SN, respondents from both countries express low levels of normative pressure to

engage in quitting (mean = 1.4).

6

Table 3. T-test results for each TPB variable across Country

Comment [E15]: Hard to see, again create tables in word with the key

information and put SPSS output in

appendix.

7

Regression Analysis of the TPB model

In order to determine the usefulness of the TPB in this context. Regression analysis

was employed to determine if any of the TPB antecedents explained intention to quit

smoking. The dependent variable is defined as the intention to quit smoking. The

independent variables are attitude, subjective norm and perceived behavioural control.

The results (see table 4) show that the TPB antecedents explain approximately 34% of the

variation in intention to quit smoking. However neither Attitude nor SN are significant

factors in explaining intention (ps > .05). Only self-efficacy explains intention to quit (p

< .05) with respondents who have stronger belief in their ability to quit smoking having a

higher intention to do so. Thus the TPB and TRA models have not been shown to be

powerful in explaining intention in this behavioural context. An additional hierarchical

regression analysis including country as an additional predictor variable shows that

country adds to the explanation of intention (p < .05) with an increase in r-squared from

Comment [E16]: Not a complete sentence. When you start with “In order to ….” You should have a second part that

goes something like “… one

should/must ….”

Comment [l17]: (neither one) is

Comment [E18]: Where is the output for this?

8

.34 to .36 (Sig. F change p < .05) This indicates that the TPB should be modelled

separately for each country.

Table 4. Regression results for TPB model

Discussion

The research results give some helpful insights how to explain the behavior of smokers to

quit. The data acquired can be useful input for organizations which develop anti-smoking

campaigns in the two countries. The fact that perceived behavioral control is the strongest

influencer of a smoker’s intention to quit is useful information that enables marketing

managers to implement suitable strategies, such as promotional programs which support

smokers to know and develop their motivation and will-power regarding quitting. An

important finding for social marketers is that respondents from Greece hold positive

Comment [E19]: In this case, you should conduct and report this additional

analysis.

Part correlations could be included and

explained, also what about beta values?????

Also demographic factors could be included

in the model.

Comment [E20]: No mention of the predictive ability of the TPB in relation to

other studies in this area or with regards to

meta-analyses on the TPB. Wherever possible, it is useful to relate your

findings to that reported in the literature

you referenced earlier.

Comment [E21]: Should use UK English spelling consistently.

Comment [l22]: Again US spelling.

Comment [E23]: Again US spelling.

9

attitudes towards smoking, and these positive attitudes might inhibit their intentions to

quit smoking.

There are various factors that limit the validity of the research results. Firstly solely the

intention to quit was measured, but not the actual behaviour. This might limit the

effectiveness of non-smoking strategies based on the results, since no evidence for the

translation of intention into the actual action exists. Furthermore other variables that were

identified to influence on the intention to quit, such as habit (Godin et al., 1992) and past

experience (Ajzen, 1985) were not included into the research.

Regarding the validity of the findings there are some limitation within the research

methodology. The very brief questionnaire cannot catch all dimensions of TPB model.

Additionally the subjective norm was only measured on a one item question. There might

have been higher correlation between the subjective norm and the intention to quit, as

identified by previous researchers, if the question would have been asked with a multi-

item scale. The comparability across the countries is limited due to some significant

differences within the sample profile by country in terms of unequal distribution and

structure. Apart from the limitations of the research the TPB model itself has some

restrictions. First of all demographics, personality-related and cultural factors are not

taken into account, although they also shape the behaviour of people. Additionally, the

model assumes rationality of people, unconscious decision making is not considered

(Sharma, 2007).

Overall the TPB has performed poorly in explaining intentions to quit smoking in Greece

and in Spain.

References

References should be included

Also useful to include an appendix (say to give a clear definition of the key concepts

such as self-brand connection, perceived value, etc. - reference to extant literature will

be good too.)

Comment [E24]: What does this mean??? When you make a statement, you

need to convey clearly and if necessary provide a justification too to substantiate

your statement/claim.

Comment [E25]: Watch out making this generalization as the analysis was not

conducted separately for each country.

Also, try to avoid short paragraphs (with

less than four sentences).