Despite differences in extent of engagement of users, original tweets and retweets to scientific publications are considered as equal events. Current research investigates quantifiable differences between tweets and retweets from an altmetric point of view. Twitter users, text, and media content of 371 randomly selected tweets and retweets linking to scientific articles published on PLoS ONE were manually categorized. Results for power of proportion differences indicated that while academic and personal accounts produced majority of original tweets (35% and 27%, respectively), they posted significantly larger proportion of retweets (43% and 41,5%). Bots and Business accounts, however, had both posted significantly more original tweets (20% and 6.5%) than retweets (2% and 3%). Natural communication sentences prevailed in retweets (79% vs. 42%) and images were found three times more in retweets (55.5%) than original tweets (18%). Overall, the findings suggest that the attention scientific articles receive on Twitter may have more to do with the inclusion of visual content in the tweets and human interaction, rather than the significance of or genuine interest towards the research results.