Assumptions tested

The assumptions tested for this study included normality and variance homogeneity tests. The data was confirmed to be normally distributed at 5% level of significance (Kolmogorov-Smirnov p =.20) hence the assumption achieved. The Levene’s test based on mean resulted in a p-value greater than 5% level of significance (p=.25) showing that the variances were equal and hence the variance homogeneity assumption was met. Achieving the two assumptions allowed for the use of parametric two-way ANOVA test.

Having confirmed the parametric assumptions, the next step involved reporting descriptive and inferential with respect to the three variables of interest (smoking, onset age and alcohol consumption). The main purpose of the descriptive statistics was to develop insight on variable of interest and check whether the response variable is appropriate for a parametric test. These are presented below.

Summary statistics test was performed for the alcohol and smoking habits. These included mean, standard deviation, minimum, maximum and skewness as shown below.

Table 1. Descriptive statistics

As displayed in above the average smoking score (M=5.93, SD = 1.86) was above the middle value with minimum maximum smoking scores of 1.77 and 9.97 respectively. The variable was slightly skewed to the left (-.14) showing that most smoking scores were high. The skewness value was however small that normality could still be assumed. Confirming the statement involved the use of the Kolmogorov-Smirnov and Shapiro Wilk tests resulted in p-values greater than 5% level of significance thus confirming the data to be normally distributed and this is displayed in table 2 below.

Table 2. Normality test

This normality was further confirmed through histogram visualization and this is also displayed normality as shown below.

Fig 1. Histogram for smoking behavior embracing normality

Having confirmed normality, it is appropriate the use of parametric ANOVA test with the data.

Next, the smoking scores were compared based on alcohol consumption and smoking onset age. The corresponding results are displayed in table 3 below.

Table 3. Contingency table

From the table, the highest smoking scores (M=7.36, SD=1,37) was recorded for the moderate to high level alcohol consumers of 20 years or more followed by none to low level alcohol consumers of 20 years and above (M=6.46, SD=1.31) showing that onset age is a factor influencing smoking descriptively. For both levels of alcohol consumption, the lowest scores of smoking characteristics were recorded for those between 0 to 19 years’ onset age group. As well, the older category recorded the largest score in total (M=6.92, SD=1.39) compared to the 0 to 19 years’ onset age group (M=5.23, SD=1.83).

Results

Two-Way ANOVA: Relationship between alcohol consumption, and onset age and smoking behavior

The first output of the two-way ANOVA test was that of Levene’s test of variance homogeneity and this is displayed below.

Table 4. Variance homogeneity

From the results, the smoking behavior variable displayed variance homogeneity based on mean, and median resulting in p-values greater than 5% level of significance. It is therefore clear that the variable fulfilled all the necessary assumptions of a parametric Two-Way ANOVA.

Thereafter, the within subject effect of the two-way ANOVA test was reported and this is displayed in table 5.

Table 5. Two-Way ANOVA

From the test, as shown, alcohol had no significant impact on the smoking habits of respondents (F (1, 76) = .22, p = .65) hence its respective null hypothesis is rejected and a conclusion made that alcohol has no role in a person’s smoking habit. In contrast, the age category at which one started smoking played a significant role in the smoking habits of individuals (F (1, 76) = 21.91, p = .00) thus its respective null hypothesis is accepted. Finally, there was an interaction effect between the smoking onset age and alcohol consumption on smoking habits of a person (F (1, 76) = 8.76, p = .004) hence its respective hypothesis is accepted.

Conclusion

In this study, it was sought to establish the influence of alcohol consumption status and onset age of smoking on the smoking habits of an individual. From the two-way ANOVA results, onset age of smoking played a significant role in smoking habits of an individual. Therefore, the hypothesis statement that a person’s onset age of smoking plays a significant role in their smoking habits is accepted as true. In contrast, the level of alcohol consumption was not a substantial influencer of a person’s smoking habit. Hence, the hypothesis statement that alcohol consumption habit plays a vital role in a person’s smoking habit is rejected as false. Similar to the onset age of smoking influence on smoking habits, there existed a joint substantial influence of onset smoking age and alcohol consumption on smoking habits and thus the respective hypothetical statement that there is a joint substantial effect of onset age and alcohol consumption is accepted as true. Comparing this study results to past works of literature, then there are both similarities and differences. To start with, Beard (2017) established that the onset age of cigarette consumption was key to establishing one’s smoking habits, a finding that’s similar to the one of this study. Particularly, Jessica et al. (2020) established that smoking characteristics were more common among the older and those who started smoking at an elderly onset age which further aligns with the findings in this study. This, however, is depended upon other confounding factors such as emotional condition at the onset age as well as the reason behind the choice to smoke. Considering the general age effect, then its anticipated by Marita et al. (2017) that most smokers are older, a factor associated with high level of exposure to smoking as well as many psychological problems suffered by people as they age. In a contrasting study conclusion, Morell et al. (2011) while studying high intensity binge drinking relationship with smoking of cigarettes established that alcohol plays a key role in smoking of cigarettes. Specifically, people with high alcohol consumption habit are more of smokers than less consumers. This is fueled by the fact that alcohol and cigarettes are sold at the same place in most instances and that the two stimulate towards the use of each other. Equally, Beard et al. (2017) established that most alcohol takers also smoke either cigarette or other substances as bhang, a finding that contradicted the one from this study. While seeking to establish if the onset smoking age and alcohol consumption plays a vital role in a person’s smoking habit, Jessica et al. (2020) found a similar result to that of this study where the two variables portrayed significance in their joint influence of smoking behavior in a person. The study results revealed that while many studies are conducted on the factors influencing smoking in persons, there is yet to be established generalizability as contradicting results are in some variables are common. Nevertheless, using most frequent conclusions from other works of literature and this, it is important to state that alcohol consumption and onset age of smoking are vital impactors of smoking habits in a person.

The study was limited to a small size and this likely influenced the study results: effect size and significance. As well, there is an error of variable omission as there are many significant influencers of smoking habits not included such as psychological stress and economic backgrounds. This is a likely cause of confounding that might have affected the significance of alcohol consumption variable. Therefore, I suggest a further study that will take into consideration a larger sample size, say 500, as this will improve the effect size and study power. I also suggest the inclusion of additional influencers of smoking to minimize the problem of confounding and improve the overall significance of the study variables.

In conclusion, it is clear from the study that smoking behavior of individuals is affected by onset smoking age and the interaction between onset smoking age and alcohol consumption. Nevertheless, alcohol consumption plays no substantial role in smoking. The study results can be applied along other works of literature to address the generalization problem associated with the factors influencing smoking habits. It can also be used for the purposes of study in both class and scientific research centers to arrive at a meaningful conclusion on the impact of onset smoking age and alcohol consumption on smoking habits by a person.

References

Beard, E. W. (2017). Association between smoking and alcohol-related behaviours: a time-series analysis of population trends in England. Addiction (Abingdon, England), 112, 1832-1841. doi:10.1111/add.13887

Fatma, Ç. S. (2018). Smoking Behavior and age. European Respiratory Journal, 52(62).

Jessica et al. (2020). TrendsintheAgeofCigaretteSmokingInitiationAmongYoungAdults intheUSFrom2002to2018. JAMA, 3(10). doi:10.1001/jamanetworkopen.2020.19022

Wang, Q. (2020). High-intensity binge drinking is associated with cigarette smoking and e-cigarette use among US adults aged 40–64 years: Findings from the 2017 BRFSS survey. Tobacco Induced Diseases, 18(54). doi:10.18332/tid/122603

Marita, H., Selma, C. L., & David, P. T. (2017). Incentives for preventing smoking in children and adolescents. Cochrane Database of Systematic Reviews (6).

Morrell, H. E., Song, A. V., & Halpern-Felsher, B. L. (2011). Earlier age of smoking initiation may not predict heavier cigarette consumption in later adolescence. Prevention science: the official journal of the Society for Prevention Research, 12(3), 247-254.

Beard E, West R, Michie S, Brown J. Association between smoking and alcohol-related behaviors: a time-series analysis of population trends in England. Addiction. 2017 Oct;112(10):1832-1841