Due to differences in cognitive and social stimuli affecting how an individual completes a questionnaire, survey researchers must deal with the consequences of mode administration differences. Mode choice affects data quality, measurement error, and estimation, potentially having implications on inference. Modes of data collection change due to lower cost, convenience, and representativeness of a sampling frame. Data collected from repeated surveys conducted over time may have incorporated different data collection modes. These changes are often associated with an inability to detect changing opinions due to possible confounding bias of mode effect. From 2006 to 2014, a biennial survey has been conducted of Oregon residents assessing satisfaction with highways and determining the most important issues facing Oregon transportation. A series of experiments were conducted on the general population to determine the impact of mode on survey results and to examine the demographics of respondents across each of these modes over time. All surveys were based on a stratified random sample of Oregon residents. For all years, mail and mixed-mode (Web followed by mail) were used to compare results. The Oregon household contact information used address-based sampling with addresses obtained from the US Delivery Sequence File. In 2006 and 2008, a random digit dialing telephone mode was also used. We implement a design based modeling perspective to estimate mode effects through time after accounting for demographics covariates such as gender, age, income, education attainment, and race. We also present a simulation study to assess the sensitivity of our proposed bias correction methodology.