Advice for automated text analysis
Budget: $10 – $30 USD
I am conducting a research project for which I need to analyze a series of texts with the “Brysbaert Concreteness Index” dictionary. This dictionary is preinstalled as “Brysbaert et al. Concreteness Norms” on this free software: https://www.ryanboyd.io/software/riot/download/
My problem is that I can’t seem to work out how to conduct the analysis for several texts at once. The attached excel contains transcripts for 895 startup pitches in column 3. I want all of them analyzed so I get the “Brysbaert Concreteness Index” scores for each of them in the same row in the following columns.
The deliverable is that you tell me how to do this. Doing it manually isn’t an option because I need to run the software on another 118.000 pitches afterwards.
Thanks in advance!
Sidenote: Inspiration for this analysis is this research article by Huang et al (2020). I want the result to be just like it.
Huang, L., Joshi, P., Wakslak, C., & Wu, A. (2020). Sizing up entrepreneurial potential: Gender differences in communication and investor perceptions of long-term growth and scalability. Academy of Management Journal:
Brysbaert Concreteness Index. We measured abstraction using the Brysbaert Concreteness Index (BCI). The BCI relies on abstraction norms that were created for 40,000 commonly used word lemmas in contemporary English (Brysbaert, Warriner, & Kuperman, 2014). Prior research on Construal-Level Theory used BCI to index speech abstraction of both shorter (Snefjella & Kuperman, 2015) and longer (Bhatia & Walasek, 2016) texts. These concreteness norms emerged from data collected on 4,000 participants who rated the concreteness of different words on a scale ranging from 1 (abstract or language based) to 5 (concrete or experience based). Brysbaert et al. (2014) defined concrete words as referring to “something that exists in reality, you can have immediate experience of it through your senses (smelling, tasting, touching, hearing seeing) and the actions you do.” In contrast to concrete words, abstract words are those that are more difficult to visualize and cannot be experienced physically, e.g., justice, morality, or strategic vision. We computed a BCI score for each pitch
based on words in the Brysbaert dictionary, calculating both a mean estimate of the overall concreteness of the language used in the pitch as well as an estimate of the overall variability in linguistic concreteness in the particular pitch2. Examples of pitches that were rated as abstract or concrete are shown in Appendix A.
[…]
We used the Recursive Inspection of Text (RIOT) scan to generate the Brysbaert Concreteness Index (Boyd, 2016). The RIOT scan is a freely available software tool for calculating indices of language from text files. More information on the RIOT scan can be obtained here: https://riot.ryanb.cc/. Like other researchers using this approach, the RIOT scan methodology does not account for bigrams in the Brysbaert dictionary (Bhatia & Walasek, 2016; Snefjella & Kuperman, 2015).
My problem is that I can’t seem to work out how to conduct the analysis for several texts at once. The attached excel contains transcripts for 895 startup pitches in column 3. I want all of them analyzed so I get the “Brysbaert Concreteness Index” scores for each of them in the same row in the following columns.
The deliverable is that you tell me how to do this. Doing it manually isn’t an option because I need to run the software on another 118.000 pitches afterwards.
Thanks in advance!
Sidenote: Inspiration for this analysis is this research article by Huang et al (2020). I want the result to be just like it.
Huang, L., Joshi, P., Wakslak, C., & Wu, A. (2020). Sizing up entrepreneurial potential: Gender differences in communication and investor perceptions of long-term growth and scalability. Academy of Management Journal:
Brysbaert Concreteness Index. We measured abstraction using the Brysbaert Concreteness Index (BCI). The BCI relies on abstraction norms that were created for 40,000 commonly used word lemmas in contemporary English (Brysbaert, Warriner, & Kuperman, 2014). Prior research on Construal-Level Theory used BCI to index speech abstraction of both shorter (Snefjella & Kuperman, 2015) and longer (Bhatia & Walasek, 2016) texts. These concreteness norms emerged from data collected on 4,000 participants who rated the concreteness of different words on a scale ranging from 1 (abstract or language based) to 5 (concrete or experience based). Brysbaert et al. (2014) defined concrete words as referring to “something that exists in reality, you can have immediate experience of it through your senses (smelling, tasting, touching, hearing seeing) and the actions you do.” In contrast to concrete words, abstract words are those that are more difficult to visualize and cannot be experienced physically, e.g., justice, morality, or strategic vision. We computed a BCI score for each pitch
based on words in the Brysbaert dictionary, calculating both a mean estimate of the overall concreteness of the language used in the pitch as well as an estimate of the overall variability in linguistic concreteness in the particular pitch2. Examples of pitches that were rated as abstract or concrete are shown in Appendix A.
[…]
We used the Recursive Inspection of Text (RIOT) scan to generate the Brysbaert Concreteness Index (Boyd, 2016). The RIOT scan is a freely available software tool for calculating indices of language from text files. More information on the RIOT scan can be obtained here: https://riot.ryanb.cc/. Like other researchers using this approach, the RIOT scan methodology does not account for bigrams in the Brysbaert dictionary (Bhatia & Walasek, 2016; Snefjella & Kuperman, 2015).