OBJECTIVES: Mainstream of psychological researches usually utilizes large samples, cross-sectional studies and aggregate frames to analyze data and interpret them. However, the limitations are that the specific features of individuals are not easy to be revealed and the time effect has been ignored. This study explores both of the trait-like and state-like properties including global self-esteem, happiness and the quality of life. By using time series analysis which can examine individual, longitudinal and non-aggregate data, the properties of psychological measures can be investigated.  METHODS: Ten college students (mean age=20.6 year, SD=2.18) and four adults (mean age=30.27 year, SD=1.23) participated in this study. Each subject completed six 10-cm visual analogue scale items, once a day for 2 months. These items measure subject’s global self-esteem, happiness, the quality of life, positive life events, negative life events as well as random error (i.e., Participant was asked to make a mark on the center of a line). Time series analysis, including autoregressive and moving average modeling procedures, was used to examine the time dependency (i.e., more trait-like) for each item. RESULTS: For the college sample, 80% of autocorrelation and partial-autocorrelation coefficients were not significant across time lag. This result doesn’t fully support the existence of time dependency for each item. On the contrary, 55% of autocorrelation and partial-autocorrelation coefficients were statistically significance for the adult sample. Both samples showed significant correlations among psychological measures (self-esteem, happiness and the quality of life) and life events (positive and negative). CONCLUSIONS: In conclusion, the results support that self-esteem, happiness and quality of life may contain both trait-like and state-like properties, and the time dependency of psychological measures is more stable especially in adults.