Saturday, March 28, 2009

True/Quasi Experiments and Quant Descriptive Studies

Lauer and Asher:
True Experiment:
Special type of experimental design in which the researcher actively intervenes systematically to change or withhold or apply treatement (or several) to determine what its effect is on criterion variables.
It's istinctive feature is the use of randomization by which subjects are allocated to treatment and control groups
When researchers cannot randomize groups - QUASI experiments are useful
- They enable cause / effect claims
- They have subject, treatments, an criterios - the same as true experiments
- No randomization of subjects but use of already intact groups
- Must have one pretest or prior set of observations on the subjects to determine in groups are qual or not on specific variavles tested in pretest
- Requires research desgin hypothesis to account for ineffective treatments and threats to internal validity.
Two types of Qausi experiments -
Weak - initially unequal groups
Strong - have equal groups - based on those variables tested in the pretest
Qauntitative Descriptive Studies:
Goes beyond case studies and ethnographies (how? it uses statistics to evaluate the data collected) to isolate sysematically the most important variables developed by these studies - define them further - quantify them - and interralate them -
Researchers -
- observe certain defines and quantifies variables for a large number of subjects over short time or short intervals of time
- Correlate variables by statistical means - Chi square / Phi coefficient / t-tests / multivariate analysis
- Descriptive - b/c there are no control groups
- Not able to make cause - effect claims
Research / Purpose
To determine which barriers are perceived to be the most frequently encountered that may affect listening effectiveness among business college students and to expand the Watson and Smeltzer study.
Questions :
- What specific listening barriers do stuents perceive as the most frequent
- What are the listening barriers factors among stuents
- How do the listening barrier factore differ based on selected stuent demographic variables.
Subject Selection:
All attended a major southwest university
81 percent were majors in the college of business : majors were General business and Finance
Males between 19 - 21
Data Collection:
Given instructions on various listening concepts
attended the business communication lecture then attended breakout sections on another day
10 of the 33 breakout sections were selected through random sampling
- Each student complete questionaire containing barriers to effective listening
Asked to indicate their perception of the frequency of each barrier that may inhibit or impede listening
Data Analysis:
279 questionairs were collected and used in analysis - statistical analysis system (SAS) was used to analyze the data
5 largest barriers - lack of interest , daydreaming, distraction, concentrating on speaker's mannerisms, deatouring

Purpose / Research
Investigates how new subjects in science and technology are represente in popular media. - Examines initial construction of nanoscale science and technology in written popular media.
Subject selection:
Public media that discusses nanoscalescience and technology
More specifically articles found in the proquest database
Data Collection:
Lit review
Data Analysis:
Divided articles up into newpaper / journal / etc
Examine propositional content of each article and grammatical structure
Developed interpretive categories of theme, rheme,topic, representation
The clause - was unit of analysis - all other categories influence the meaning and structure of the clause
process of presentice tenical info for general audiences is enables by combining social and technical approaches

Notarantonia and Cohen Quasi
Purpose / Research:
Utilizes Dominance and Openness in terms of perceived communicator style and sales effectiveness
Subject Selection:
Undergraduate business administration stuents enrolled at Bryant College. 80 in all.
41.3 - male
58.7 female
Age range - 17 - 21 y.o. - 73% 18 y.o.
92.1 % freshman

Data Collection:
four video tapes - pretested for openness/dominance categorization (175)
Subjects - completed self-report of Norton's communicator style measure then shown video tape - Questionaire

Data Analysis:
Two way - anovas(likert) - with openness and dominance as independent variables
Group answers compiled and analysed in terms of the video tape they were shown (results on pg 176)

Kroll Quasi
Research / Purpose
Explore the patterns of thought and language that characterize growth in informative writing ability across junior and senior highschool years. (198)
Subject Selection:
students from 4 schools in a small city in central Iowa and freshman at Iowa State University.
24 - 5th grade; 26 - 7th grade; 19 - 9th grade; 27 - 11th grade; 27 college freshman
All were enrolled in an English class; all had average school achievement for their ages, all native speakers (198)

Data Collection:
students told their task of writing directions about the game; instructional film shown twice; students given time to write instructions; given (pretested) m/c quiz upon completion of instructions to ensure they understood the rules of the game.(201)
Data Analysis:
Only analysed the directions of those students able to demonstrate they understood the basic rules for playing the game.
2 raters - interrater reliability - .92 for instructions / researcher rated college students response contrasted with second rater.
Analyzed based on essential / clarifying information
chi-square test / z-statistics /anova

Caroll Minimal Manual - usability testing True experiment
Research / Purpose:Emperically developed training manual for word processing functions of popular commerical office information system
Connection b/n learning purpose - training manuals -integration of learning objectives with realistic tasks for learners (75) Will this be successful?

Data Collection:
Observation (81-83)
Experiment 1 - Experiment 2 - ability to complete tasks
treatment - their manual or a standar manual
Subject Selection:
Prescreen/test - 19 in first /32 in second experiment ensured equality across groups.
Data collection:
- observation of users performing typical tasks (81)

Data Analysis:
First Experiment: Times to complete tasks were compared b/n their manual and the other
Second experiment: Anova two way scale of variance comparison b/n their manual and a standard manual with regards to task completion and preformance efficiency