Spss for windows version 20

We reduced the three color variables (L, a, and b) using principal component analysis. The first principal component (PC1) represented 63% of variance in a, b and L (eigenvalue = 1.89; 2,711 flowers); thus, PC1 is considered the flower color variable in further analysis. Note that we have additionally performed the analyses using both PC1 and PC2 (jointly explaining 95% of variation in flower color) and results remained the same (results not shown). Thus, we present results using only PC1 as a color descriptor for the sake of simplicity. Low scores of the PC1 indicate dark-orange color and high scores indicate bright-yellow color since factor coordinates of the original variables on PC1 were L (brightness) = 0.686, a (red related component) = -0.360, and b (yellow related component) = 0.631.
Although we focused on flower color, we included other phenotypic traits in the following analyses to control for their possible effects on pollination, seed predation and seed production, or their correlation with flower color. We performed Pearson correlations for the traits under study.
We followed the approach proposed by Herrera et al. [22 ] to test for natural selection driving local adaptation and the role played by the animal community. To this end, we performed the following analyses (using the SPSS for windows version 20, IBM SPSS Statistics 20).

Data were analyzed using IBM SPSS for Windows, Version 20 (IBM Corp., Armonk, NY, USA). Data were presented as mean values and standard deviations. Descriptive statistics and frequency of daily PA and risk factors of MetS were calculated for both percentage and absolute values. An independent t-test was used to examine differences in risk factors of MetS based on sedentary time, LPA and MVPA. Binary logistic regression analysis was used to identify a significant impact of PA patterns on MetS status. Multiple linear regression was used to compute the model of significant independent variables of PA patterns on risk factors of MetS. A general linear model (univariate ANOVA) was used to assess the interaction between LPA and MVPA. An α-level of 0.05 was used to determine statistical significance.

The data was analysed using IBM SPSS for Windows version 20. Categorical variables have been presented as frequencies and percentages. To examine associations between socio-demographic characteristics, food habits and the nutritional status, Chi-Square test was performed. Fischer’s exact test was used in cases where conditions for Chi-Square test were not met. P-value of < 0.05 was considered significant at 95% confidence interval.