The children were divided into four groups, two groups of five stutterers and two groups with six fluent speakers (Ambrose & Yairi, 2002.). Throughout January and May, each student had individual sessions with an experimenter who gave them one of two responses (Reynolds, 2003). One group of stutterers and one of non-stutterers were told that they should think carefully about speaking fluently and not speak if they couldn’t do so properly, otherwise they would become unintelligible stutterers …show more content…
The first was that unlike the selection of the twelve non-stutterers being random, the assignment of the groups was not. Mary Tudor, the student who conducted the study for her master thesis, split the two large groups into their smaller divisions herself (Ambrose & Yairi, 2002). For more accurate results, the stuttering and non-stuttering groups should also have been divided in half through random selection. Secondly, all the students in the study were orphans. This extraneous variable could have had an effect on their confidence, which meant that the negative treatment they received from experimenters could have affected them more than other children. Lastly, the average IQ of the participants of the study was 85.9 (Ambrose & Yairi, 2002). This low IQ is also a participant variable that could have had an effect on their speech fluency (Goldfarb, 2006). To improve the validity of the study, a large number of children from a range of backgrounds with a wider range of IQ’s should have been chosen. This would’ve ensured that the independent variable (the type of speech therapy) was the causation of the