Forward telescoping, the reporting or dating of events as being more recent than they actually were, is often observed in surveys and produces inaccurate data. We believe that some forward telescoping occurs when the question format allows people to respond without extensive retrieval of temporal information concerning the target events. We collected two types of data. The first, the type usually collected by survey researchers, involved visits to medical doctors. As is common in survey research, the actual dates of the events were not verifiable. The second type involved students' participation in laboratory research studies. Here, the actual dates were verifiable. We demonstrate that modifying the questions asked produced differences in the amount of forward telescoping in participants' responses.