The research process began with entry-level field work. I engaged with a set of 9 Hindi-Urdu speakers with varying language contact experiences. I tested their responses to sample word order stimuli, but also asked them a number of questions about language experience and attitude. The responses to the stimuli were individuals' assessments of the grammaticality of the sentences on a 1-7 Likert scale, which also served as a proof of concept for the greater experiment. The questions about linguistic demographic data were also used to construct the formal study's linguistic demographic survey and avoid any potential biases.
The stimuli were developed so that there were main experiment stimuli that tested word-order acceptability, sub-experiment stimuli that texted post-positions, and filler stimuli that were grammatical and ungrammatical to test the main experiment against. I developed these stimuli and used feedback from other Hindi-Urdu speaking linguists to adjust some of the vocabulary.
Then, I recorded the stimuli in PRAAT (a speech analysis program for phonetics). The image on the left shows an overlay of the intonational contours of OSV stimuli. The experiment utilizes audio files to elicit responses so participants only need to be able to understand Hindi-Urdu, not read, write, speak, or produce it.
I created these graphics to share on social media to explain the study and to source participants. The study was jointly hosted on Qualtrics (for the demographic survey) and on Google Firebase (for the acceptability portion).
Since the data collection portion of the experiment, we have conducted data analysis in R and analyzed the qualitative data as well. Our findings have translated into multiple abstracts, poster sessions, and papers both submitted and in progress! This study has challenged me to take a large question and use quantitative and qualitative data gathering and assessment methods in hand with prior literature to find answers and more questions that come from those answers!
