12/30/2012 Hypotheses Prepared by: Amanda J. Rockinson-Szapkiw Liberty University Hypothesis: Purpose To determine if a theory has the ability to explain, predict, or describe, you conduct experimentation and observation to test inferences derived from the theory. For example, a general inference based on experience and research may be that “community increases learning in a distance education classroom” community increases learning in a distance education classroom. In order to test one of these inferences, you need to express it in operational terms, that is formulate a hypothesis, e.g. Students’ high scores on the community scale will be highly related to good grades in the distance education course Hypothesis: Purpose Provides the researcher with a A hypothesis is an relational statement that is educated guess, and it directly testable in the study can be supported or refuted through Provides direction to the experimentation or research observation. If the observation If the Provides a framework for hypothesis is supported, reporting findings and evidence exists that conclusions of the study supports the theory. If the hypothesis is refuted, the theory is not supported. 1
12/30/2012 Hypothesis: Acceptability Characteristics (Bartos, 1992) States the expected relationship between variables, Is testable – all hypotheses need to be evaluated. Is founded in the problem statement and supported by research I f d d i h bl d d b h – literature provides a strong rationale. Aligns with the research question Stated simply and concisely as possible Hypothesis: 2 Types Research – a clear statement about the expected relationship between/among variables. Directional – specifics the direction of the expected relationship Nondirectional – does not specify the N di ti l d t if th direction of the expected relationship Null – states that no relationship exists between the variables under study Hypotheses: Application Research Question. Will there be a difference in levels of responsibility exhibited by oldest children and youngest children in a family? Research Hypothesis. There will be a statistically significant difference in the level of responsibility exhibited by oldest children in a family and that of h b d b ld h ld f l d h f youngest children in a family as measured by the Behavior Rating Profile. Null Hypothesis. There will be no statistically significant difference in the level of responsibility exhibited by oldest children in a family and that of youngest children in a family as measured by the Behavior Rating Profile. Taken from: LaFountain & Bartos 2
12/30/2012 Hypotheses : Application Research Question. Will the level of social skills of college students diagnosed AD/HD increase as a result of being involved in an interpersonal relationship group? Research Hypothesis. There will be a statistically significant increase in the level of social skills of college students diagnosed AD/HD who are involved in an interpersonal relationship group as opposed to those college students diagnosed AD/HD who are not involved in an interpersonal relationship group as measured by the FIRO Awareness Scales. >>>>> Hypotheses: Application Null Hypothesis. There will be no statistically significant increase in the level of social skills of college students diagnosed AD/HD who are involved in an interpersonal relationship group as opposed to those college students diagnosed AD/HD who are not involved in an interpersonal relationship group as measured by the FIRO interpersonal relationship group as measured by the FIRO Awareness Scales. Taken from: LaFountain & Bartos Hypothesis: Application In the previous examples the hypotheses are much more specific than the research questions. The hypothesis is written in such a way that makes it testable. We use the term “statistically significant” because a difference or increase is not enough; the variance must be at a statistical level. 3
12/30/2012 Hypothesis – How many hypotheses? Rule The number of hypotheses needed is based upon the number of variables under study the number of variables under study. Hypothesis – How many hypotheses? Rule If you are conducting research in differences between groups and have one independent variable and one dependent variable you test one null hypothesis, a main effect hypothesis. Hypothesis – How many hypotheses? Rule If you have two independent variables and one dependent variable you test three null hypotheses: two main effect null hypotheses and one interaction hypothesis, which addresses a first order interaction. 4
12/30/2012 Hypothesis – How many hypotheses? Rule If you have three independent variables and one dependent variable you test seven null hypotheses: three main effect null hypotheses, three first order interaction hypotheses, and one second order yp interaction hypothesis. The first order interactions address interactions between all possible pairs of independent variables. The second order interaction involves all three independent variables. Hypothesis: Example The following are null hypotheses for a difference question in which the independent variables are Type Distance Education Course (synchronous, asynchronous) and Gender (male, female), and the dependent variable is sense of community: H O1 : There is no statistically significant difference in sense of community among students based on type of distance education course (synchronous, asynchronous) . ( This hypothesis addresses the first IV or main effect or factor .) y ) ( yp f ff f ) H O2 : There is no statistically significant difference in sense of community among students based on gender (male, female) . ( This hypothesis addresses the second IV or main effect or factor .) H O3 : There will be no statistically significant difference in sense of community among students based on type of distance education course (synchronous, asynchronous) remains constant regardless of gender (male, female). ( This hypothesis addresses the interaction effect .) Hypothesis – How many hypotheses? Rule If you plan to examine two or more correlated dependent variables, then you need you need to pose a hypothesis for the linear combination of variables as well as each variable separately. Note that I said p y correlated here. If dependent variables are significantly correlated, they need to be analyzed together n order to control for possible Type II error. They should be considered separately if they are not correlated. 5
12/30/2012 Hypothesis: Example Let’s consider the question, What is the effect of participation in the Math 2.0 program on second grade students’ math motivation scale scores? Here the math motivation scale instrument consists of three correlated subscales (e.g. external, introjected, and identified regulation). Thus, there is one independent variable but three dependent variables. This research question requires four hypotheses, H : There is no no statistically significant difference in second grade students’ mean H 01 : There is no no statistically significant difference in second grade students mean scores on the math motivation scale based on type of math program participated in. H 02 : There is no statistically significant difference insecond grade students’ introjected mean scores on the math motivation scale based on type of math program participated in. H 03 : There is no statistically significant difference in second grade students’ identified regulation mean scores on the math motivation scale based on type of math program participated in. H 04 : There is no statistically significant difference in second grade students’ external mean scores on the math motivation scale based on type of math program participated in. Hypothesis Rule It is also important to note that control variables need to be reflected in the hypotheses. Here independent and dependent variables are discussed; however, the rules hold true when examining variables of , g interest in a correlational study. The number of hypotheses needed is based upon the number of variables under study. Application Develop Your Research and Null Hypotheses 6
12/30/2012 Hypothesis : Application Research Question Research/ Null Hypothesis Is there a difference between online students’ H1: Online students who utilize a combination perceived sense of community or inquiry (e.g. of synchronous and asynchronous CMC social presence, cognitive presence, and teacher systems for instruction and discussion as presence )who use only asynchronous CMC opposed to online students who utilize only an systems and who use a combination of both asynchronous CMC systems for instruction and asynchronous and synchronous CMC systems? discussion will have statistically significant differ in terms of the combination of social presence, cognitive presence, teaching presence, and learning outcomes. H2: Online students who utilize a combination of synchronous and asynchronous CMC systems for instruction and discussion as opposed to online students who utilize only an asynchronous CMC systems for instruction and discussion will have statistically significant different levels of social presence as measured by the COI Framework survey (Arbaugh et al., 2008). 7
Recommend
More recommend
