PART ONE: Study Description

Research has shown that eyewitnesses are susceptible to the "misinformation effect", or the phenomenon in which witnesses incorporate "misinformation" into their memory after receiving misleading information about an event. Researchers designed a study to see if an eyewitness who saw a theft incorporated such misinformation into their memories. They tested participants one at a time, but used several confederates (experimenters pretending to be unaffiliated with the study) in the design to provide misinformation.

Set-up: The experimenter recruited participants to take part in a computer study in a room with four laptops. One laptop was always left unused, two laptops were used by confederates #1 and #2, and the real participant used the remaining laptop. Participants were tasked with a point-and-click speed activity where they had to click as quickly as possible on an emoji (smiling face with sunglasses) that appeared at random on the computer screen. Fifteen minutes into the computer task, a thief (confederate #3) entered the room, walked around, and picked up the unused laptop. Confederate #1 said to confederate #2, "Is that guy supposed to be here?", to which confederate #2 replied, "If he is, he's really late." This conversation was loud enough so the real participant could overhear the exchange (This conversation was scripted and practiced well before the study began and was used for all participants. We thus ensured that the participant looked up from their own laptop to see what was happening). Confederate #3 unplugged the laptop and ran out of the room with it. Confederate #1 shouted to the experimenter, "Hey, I think that guy just stole your laptop." Although the experimenter chased the thief out of the lab, she returned a moment later saying he got away. The experimenter then called campus security (yup, more confederates!).

The participants were individually questioned about the theft by "campus security" on day one (yup, those additional confederates) in a private room and asked about what they had seen, including what the thief looked like, what he wore, how old he was, etc. The participants were asked to return the next day (day two) to speak with a police detective. The script then diverged for the experimental and control conditions.

In the experimental ("misinformation") condition, the participant and confederates #1 and #2 left together. In a whispered conversation with confederate #2, confederate #1 said, "That guy had a gun. I've never been so scared in my life." Confederate #2 simply responded with the phrase, "Yeah, it was bizarre." Both then hurried away, noting that they were late for their next class.

In the control condition, confederate #1 merely said, "I've never been so scared in my life", to which confederate #2 again responded, "Yeah, it was bizarre". There was no mention of a gun.

On day two, a police detective (actually the primary researcher) questioned the participant and asked if the participant had seen a gun. For purposes of data analysis, participants were asked how sure they were that the thief had a gun, ranging from 0 (there was definitely no gun) to 100 (there was definitely a gun).

Using this study set-up, identify which internal validity threat is most likely in each of the following statements.

Question 1

The study ran for several weeks during the semester. About a week after it started, the university announced that they would be holding training sessions for faculty and staff about how to handle situations involving a gunman on campus. This shut down the study for several days as the university needed the lab building for training. The study then resumed according to script after the training. The researchers found that those in the experimental group did not differ in their memories regarding the presence of a gun compared to those in the control condition. That is, the mean score that a gun was present was similar for the experimental group (M= 65%,SD= 11.4%) and the control group (M= 63%,SD= 13.26%)

Interactions

Diffusion

History

No Threat

Instrumentation

Mortality

Maturation

Testing

Regression to the Mean

Selection