Smart phones take psycholinguistic tests on the go

Undergraduates at research universities make up an overwhelmingly disproportionate number of participants in cognitive science experiments, such as psycholinguistic tests. This skewed sampling reduces the ability to make inferences from the sample to the general population; thus, the inability to recruit large heterogeneous samples has been a major limiting factor in cognitive science. Smart phones may provide the opportunity to tap into a larger, more heterogeneous sample. Stephane Dufau at CNRS & Aix-Marseille Universite created an app that he believes will revolutionize how data is collected in cognitive science.

“Science XL: word power test” can be downloaded from itunes for free. This app is an ios-compatible adaptation of the English Lexicon Project. Users can choose to do 50, 100 or 140 trials. Each trial consists of a fixation cross, which is followed by the stimulus, either a word or non-word. The stimulus remains displayed until a response ("yes" or "no") is chosen. Initial experiments with Science XL: word power test showed a linear relationship between data collected with the app versus data collected in a controlled lab setting. The word power test is the first step at moving data collection out of the lab and into the real world, with the overall goal of Science XL being to tap into the power of smart phones as a scientific research tool.

Smart phones are increasingly becoming an integral part of users' daily lives. A study by Google found that a large proportion of smart phone users use their phones to multitask while consuming other media. Science XL attempts to utilize the widespread use of smart phones to revolutionize data collection. While it is true that smart phones provide access to a large sample of individuals, the use of smart phones as a research tool does not eliminate the self-selection sample bias. Individuals who own and use smart phones are certainly a self-selected group. Furthermore, individuals who would download, use and email their data for the Science XL: word power test would be another level of self-selected individuals. Although Science XL does ask for some very basic information such as gender and handedness, it does not account for a variety of other factors like education or reading level that would contribute to performance on the test. It also doesn’t account for any environmental factors that could alter performance on the test.

It is interesting that smart phones are being used as research tools that assess our ability to differentiate between words and non-words. As avid text messagers know, texting often results in non-words or autocorrected words that don't make sense. Furthermore, the rise of texting is creating its own lexicon of shorthand words. How this plays into our ability to discriminate words from non-words quickly in this task remains unclear.

As usage of smart phones continues to rise, it is clear that they provide an amazing avenue for mass data collection. From consumer patterns to reading ability, we are beginning to scratch the surface of the type of information that can be easily collected from the masses through the use of smart phones. Although self-selection is still an issue, the data collected from the Science XL: word power test certainly comes from a larger and more heterogeneous sample than traditionally available.