Depression is one of the most common mental health disorders affecting people from different age groups, societies, communities, and countries. Many countries lack awareness about psychological disorders and there is a scarcity of good mental healthcare facilities available globally. Medical practitioners recognise depression by analysing the patient’s behavioural patterns like speech levels, facial expression, body language and language patterns during therapy. Previous research has shown that behavioural studies are effective means for depression recognition. To explore this relationship, the automated depression recognition dataset called Distress Analysis Interview Corpus (DAIC) was evaluated. This dataset was chosen as it consists of paralinguistic (vocal), linguistic (verbal/text) and extralinguistic (visual) features from the dyadic interviews between participants and a virtual human. Prior research has evaluated the relationship between paralinguistics, linguistics and depression but many researchers failed to analyse the relationship between personality and depression for the DAIC database. The present study explores how different paralinguistic and linguistic features and personality types differentiate between high and low levels of depression. This study was exploratory in nature and used the LIWC software for linguistic and personality analysis, Pandas software for pre-processing the audio and text data files and lastly correlational analysis using JASP software to answer the research questions. The main findings concluded that linguistic features like emotion (sad and negative), feeling and health related words are used most often by depressed people. Additionally, paralinguistic features like high pitch and breathy voice as well as the personality trait neuroticism were characteristic identifiers of depressed people. These results showed that linguistics, paralinguistics, and personality traits help in depression recognition. These research findings have the scope for broader and cross-disciplinary applications in the future. Further research and development for automated detection technologies is required in the field of behavioural studies, to enable people globally to easily access and use artificial healthcare platforms for mental health diagnosis.