Demographics
Results showed that out of the total of 328 invited participants, 206 participants covering 7 gyms participated in our study. Our study yielded a response rate of 62.8% across gyms. Participants (n= 206) included 151 males (73.3%) and 54 females (26.2%). The majority of our sample (76.1%) were between the ages of 18 and 35, 14.1% of it were below 18, and the remaining 9.8% were 36 and above.
Description of exercise type and frequency among the study sample
Our results suggest that respondents had diverse exercise types, with 63.6% of the sample reporting that they practice both aerobic and anaerobic exercise, 18.4% reporting that they practice only anaerobic exercise, and 14.6% saying that their exercise is aerobic. Across these diverse exercise types, 54.8% of the sample has been working for a duration of more than 6 months, indicating that the sample is generally experienced with athletic performance and physical activity. Additionally, 33.5% which represents the majority of our respondents work out for a duration of 30 to 60 minutes, Also, 30.9% of respondents spend 61 to 90 minutes during their workouts. Consequently, 14.9% of the participants’ report that they work out for less than 30 minutes and 19.1% of them practice for more than 90 minutes indicating that the participants work out for a good amount of time.
For most of the following analyses, participants will be distributed in two pools, those who practice aerobic exercise (n1= 161) and those who do anaerobic exercise (n2=169). The mean frequency of exercise among pool 1 (aerobic exercise) is 4.31 days/week (SE=0.334) and that among pool 2 (anaerobic exercise) is 4.40 days/week (SE= 0.143).
The majority of our participants from both pools, anaerobic exercise and aerobic exercise, report that they exercise for 5 to 6 days/week. As shown in Figure 1, 30.7 to 31.8% practice anaerobic exercise 5 to 6 days respectively. Additionally, 29.2 to 31.2% practice aerobic exercise 5 to 6 days respectively (Fig. 2).
Individual beliefs about the effect of caffeine
Caffeine consumption was popular in our sample, with 92.2% indicating that they use a certain caffeinated product, most commonly coffee, soft drinks, and tea. The caffeine expectancy questionnaire included statements related to the influence of caffeine and asked participants to rate their level of of agreement with the statements. Table 1 shows that participants agreed the most with the statement “caffeine picks me up when I am feeling tired” (Mean= 3.55, SE= 0.252) and disagreed the most with the statement “I am easily stressed after having caffeine” (Mean= 2.41, SE= 0.172) suggesting that the sample is experienced with similar knowledge about the effect of caffeine.
Reliability test
The reliability coefficients were conducted for the caffeine questions (independent variables) that consisted of 16 items. The resulting Cronbach’s alpha was equal to 0.901 which is above 0.7 demonstrating a good level of reliability. Scores demonstrated an acceptable level of reliability. Thus, caffeine questions are consistent, all reliably measure the same independent variable related to individual beliefs of caffeine consumption.
Factor analysis
Since our 16 items related to individual beliefs about the effect of caffeine reflect different dimensions underlying caffeine beliefs, we proceeded with conducting a factor analysis. First, to make sure that factor analysis is applicable, the Kaiser-Meyer-Olkin (KMO) test and Bartlett’s test for sphericity were conducted on the 16 items. The KMO test assesses how suited the data is for factor analysis and it is a measure of the proportion of variance among items that might have common variance. The larger the proportion, the more suited the data is for factor analysis. KMO values between 0.8 and 1 indicate that the sampling is adequate and factor analysis is meritorious or marvelous [15]. We found that this proportion is 0.887. The Bartlett’s test was significant (p value <0.0001) indicating that the items were correlated and therefore suitable for data dimension reduction and hence factor analysis may be useful. Thus, the results of both KMO and Bartlett’s tests agreed that a factor analysis can be conducted.
The output in (Fig. 3) mainly shows the presence of the first, second and third factors as influential points with well-presented angles. Therefore, we conclude that the data may be reduced to three factors only which are factor 1 related to physical performance, factor 2 related to work durability, and factor 3 related to productivity.
In the following Table 2, we show the results obtained from the factor analysis with the percentage of variances and the eigenvalues obtained for all factors. The first four factors have a total of 67.8% cumulative variance. The first three factors were included, and we excluded the fourth one since it reflects a small number of items. In addition, items underlying this factor were detected as outliers.
The eigenvectors for the first three factors, are presented in the following Table 3. All the values of the first vector are positive, hence the vectors were sorted by the second and then by the third vectors. Based on the eigenvectors values, four groups of variables were identified. The first group of items include: 3 (caffeine improves my athletic performance), 6 (workouts are better after having caffeine), 7 (caffeine increases my motivation to work), 13 (I would be unable to function without caffeine) and 15 (I can exercise longer if I have caffeine).
The second group of items with values of vectors 2 and 3 as negative include: 1 (caffeine picks me up when I’m feeling tired), 5 (caffeine improves my mood), 11 (caffeine improves my concentration), 12 (caffeine helps me work over a long period of time), 14 (caffeine improves my attention) and 16 (caffeine makes me feel more energetic). The third group of items include 4 (I feel less sleepy after having caffeine), 8 (caffeine at any time throws off my sleep) and 10 (caffeine makes me feel more alert). The fourth group of items with all values positive include 2 (I am easily stressed after having caffeine) and 9 (caffeine makes me feel nervous).
Correlation analyses
These were carried out to show the association of the three dependent variables; aerobic exercise frequency, aerobic exercise duration, and anaerobic exercise frequency with the three independent variables based on the caffeine questions; physical performance, work durability, and productivity.
Results in Table 4 revealed the correlation between the variables. Aerobic variables were not correlated with any of the independent variables (all p-values were not significant) suggesting that caffeine has no association with aerobic activities when perceived as a physical performance enhancer, work durability enhancer, or productivity enhancer. However anaerobic exercise frequency has a positive significant correlation with physical performance (p=0.05) and work durability (p=0.032), but doesn’t affect productivity. Our results showed that consuming caffeine increases physical performance and work durability when engaging in anaerobic activities, while it has no effect when engaging in aerobic activities. However, consuming caffeine does not have any effect on productivity while performing aerobic and anaerobic exercises.