Welcome this is. Amanda raccoons and zap goo and in this tutorial I will be discussing with you. Hypotheses and hypothesis **research hypotheses are, research hypothesis are** testing we will look at a definition for hypotheses and look how to write up different hypotheses we will then talk about hypothesis **research hypotheses are, research hypothesis are** testing. And we’ll look at the different steps including stating the null hypothesis **research hypotheses are, research hypothesis are** choosing a statistical significance level carrying out the appropriate appropriate statistic and making a decision regarding that statistic will also talk about type 1 and type 2 errors. Let’s start by talking about a hypothesis **research hypotheses are, research hypothesis are** and defining it in order to establish whether a theory has the capability to describe to explain or to predict a behavior in the world world. It’s necessary to construct. Tests of inference arising from that theory inferences are generally expressed at a conceptual level for example we might say experienced. Educators are better teachers however in order to actually perform tests it becomes necessary for us to express these in operational terms or express them as hypotheses for example we may take the inference experience education. Educators are better teachers and put it in operational terms and say that the proportion of educators who are evaluated as outstanding by a panel of experts is highly related to their number of years teaching now hypothesis **research hypotheses are, research hypothesis are**. Then is an educated guess it can be supported or refuted through or by experimentation or observation another way of defining hypothesis **research hypotheses are, research hypothesis are** is a hypothesis **research hypotheses are, research hypothesis are** is a statement about a future event or an event the outcome of which is unknown at the time of the prediction set forth in such a way that it can either be rejected or not rejected now. If a hypothesis **research hypotheses are, research hypothesis are** is supported evidence exists that supports the theory if the hypothesis **research hypotheses are, research hypothesis are** is refuted then. This theory is not supported as finding support hypotheses related to a theory the theory gains trustworthiness for example. Noel’s Adult Learning Theory has been tested and the inference. Adult learners are independent and self-directed has been supported through multiple observations through multiple experience experiments as such adult educators.

Trust this theory and they therefore apply it. Some adult educators will ask students even to formulate their own learning objectives during the first week of the course and encourage independence which encourages independent and adult students control. Then their own learning so a hypothesis **research hypotheses are, research hypothesis are** is a statement about a future event or an event the outcome of which is unknown at the time that is set forth in such a way that can be rejected or not rejected. And whenever there’s support for the theory around a support for a hypothesis **research hypotheses are, research hypothesis are** around a theory that theory gains trustworthiness and when it’s considered trustworthy then practitioners often apply it now a hypothesis **research hypotheses are, research hypothesis are**. Remember a statement about a future event is stated in two ways for a research study it stated as a null and an alternative or research hypothesis **research hypotheses are, research hypothesis are** a null hypothesis **research hypotheses are, research hypothesis are** states that there’s no relationship or no difference that exists between variables under study whereas an alternative or research hypothesis **research hypotheses are, research hypothesis are** states that there is a clear relationship or different among or between variables. Now every research question has at least one null hypothesis **research hypotheses are, research hypothesis are** and one research hypothesis **research hypotheses are, research hypothesis are**. The research hypothesis **research hypotheses are, research hypothesis are** can be stated in one of two ways it can be stated as a directional hypothesis **research hypotheses are, research hypothesis are** or a non directional hypothesis **research hypotheses are, research hypothesis are**. A directional hypothesis **research hypotheses are, research hypothesis are** specifies the direction of an expected relationship and it indicates that the researcher plans to do a one-tailed analysis or test a non directional high research hypothesis **research hypotheses are, research hypothesis are** does not specify the direction the expected relationship and therefore indicates that the researcher plans to do a two-tailed test for example a directional hypothesis **research hypotheses are, research hypothesis are** may state that a researcher expects a positive relationship between two variables or an increase in some students scores. From pretest to post-test where is a non directional hypothesis **research hypotheses are, research hypothesis are** would state that the researcher expects that there’s a relationship between variables or that a difference will exist in test scores between the pretest and post-test now that we understand what a hypothesis **research hypotheses are, research hypothesis are** is.

There’s two more related terms that are important to understand. The first is a statistical hypothesis **research hypotheses are, research hypothesis are**. A statistical hypothesis **research hypotheses are, research hypothesis are** is a claim about a population that can be tested with data obtained from a sample that is only a portion of the population. Oftentimes researchers want to understand the population. But they only have access to a small portion of it a sample of it so oftentimes what happens is is that researchers collect data from the sample and do statistical tests the second term that it’s important for us to understand is a hypothesis **research hypotheses are, research hypothesis are** test sometimes also called statistical significance tests the hypothesis **research hypotheses are, research hypothesis are** test is an inferential statistic procedure in which the researcher seeks to determine. How likely it is. The results of a study are due to chance in other words the researchers trying to determine whether or not to attribute the results observed either sampling error or the actual relationship or difference between the variables being studied now when I say sampling area error what I mean is is chance the possibility that chance affected the Co variation among variables in the sample statistic in hypothesis **research hypotheses are, research hypothesis are** testing the researcher decides whether or not to reject or fail to reject the null hypothesis **research hypotheses are, research hypothesis are**. Now we’re going to take an in-depth look at hypothesis **research hypotheses are, research hypothesis are** testing in this tutorial. However before we go on to discuss hypothesis **research hypotheses are, research hypothesis are** testing let me stop for a moment to review the difference between a population and a sample. It’s really important that you understand the differences between these terms as well as the notations used for each first of all. Let’s talk about population. The population is all the members of a group that a researcher plans to focus on. It’s the focus of the research. Sometimes populations are very large and the researcher can’t collect or doesn’t want to collect data from everyone in the population so the data is collected or information is collected from the sample and a sample is a smaller group.

That represents the population the overall purpose. Remember of research is the generalization of results to a population from a smaller sample. That’s taken from that population. We generally assume that our sample of as you can see n is randomly selected from our population of n and that the population is normally distributed. Now these assumptions need to be checked in any design. Even when samples are random. And sometimes they aren’t. We cannot guarantee the sample characteristics will be identical to the population characteristics and this discrepancy is what we call sampling error now all hypothesis **research hypotheses are, research hypothesis are** tests have unavoidable quantifiable risks of making the wrong conclusion statistical tests in other words always involve type 1 or type 2 risks or errors which we’ll discuss later on however here. It’s important to understand that research is done with a sample taken from a population now while we’re talking about samples versus populations something else. I want to note here. Is that the scientific notations for mean and standard deviation are represented with different symbols. So when you see different statistical formulas with these different symbols. You can note whether or not this. It’s the statistical formula for the population or for the sample so here. I’ve noted those on the slide. You can see the scientific notation for population and sample for the mean and then also for the standard deviation so now that we understand the difference between populations and samples. Let’s go ahead and move on to hypothesis **research hypotheses are, research hypothesis are** testing. There are four primary steps to hypothesis **research hypotheses are, research hypothesis are** testing. First of all the researcher states the null in the research hypothesis **research hypotheses are, research hypothesis are** or hypotheses they choose a statistical significance level. They carry out the appropriate statistical procedure and then finally make a decision regarding the hypothesis **research hypotheses are, research hypothesis are**.

Let’s take a closer look at the first step stating the null and research hypothesis **research hypotheses are, research hypothesis are** or hypotheses as I said the goal of hypothesis **research hypotheses are, research hypothesis are**. Testing is to see if there is sufficient statistical evidence to reject a presumed null hypothesis **research hypotheses are, research hypothesis are** in favor of the alternative hypothesis **research hypotheses are, research hypothesis are** or research hypothesis **research hypotheses are, research hypothesis are**. Therefore the purpose of any inferential statistic is to determine if there’s sufficient evidence to reject the null so the very first step of hypothesis **research hypotheses are, research hypothesis are**. Testing is to state what we’re studying it’s to State or specify the null hypothesis **research hypotheses are, research hypothesis are** and the alternative hypothesis **research hypotheses are, research hypothesis are**. It’s important to note that these hypotheses are derived from the research question which are ultimately derived from the literature so the literature drives the research question and then the research question then guides the statement of the both the null hypothesis **research hypotheses are, research hypothesis are** and the research hypothesis **research hypotheses are, research hypothesis are**. These hypotheses can be written in either words or in symbols for the most part for social sciences and social science research. Words are expected however they can also be stated using symbols. And we’re going to take a look at that here. You can see how the null hypothesis **research hypotheses are, research hypothesis are** as well as the alternative hypothesis **research hypotheses are, research hypothesis are** can be stated using either words or symbols. Let’s take a closer look first at the null hypothesis **research hypotheses are, research hypothesis are** remember that the null hypothesis **research hypotheses are, research hypothesis are** assumes or states that there’s no statistically significant difference or relationship between certain population values so if a researcher was writing the null hypothesis **research hypotheses are, research hypothesis are** in using words he or she may use terms or words such as there will be no effect or no difference. The researcher in conducting a correlational study may say there’s no relationship now. I’ll bring attention here to the symbols. Used for the null hypothesis **research hypotheses are, research hypothesis are**. Here you can see that symbols used for different are different than that used for relationship for relationship. R or Rho is the symbol used not the P and Rho equals the or represents the relationship or strength of the relationship.

And here you can see. Rho is equal to zero meaning. There’ll be no relationship for the alternative hypothesis **research hypotheses are, research hypothesis are** which assumes or states that there will be a statistically significant difference or relationship between these population values being studied the researcher uses terms such as there will be a statistically significant difference or a statistically significant relationship if he or she’s writing the alternative hypothesis **research hypotheses are, research hypothesis are** using words. And then here you can also see the symbols used for the null or for the alternative hypothesis **research hypotheses are, research hypothesis are**. Now you remember that a research or alternative hypothesis **research hypotheses are, research hypothesis are** can either be proposed directionally or non directionally and this then determines the type of terminology used or symbols used when stating the researcher alternative hypothesis **research hypotheses are, research hypothesis are** if the researcher desires to state a general hypothesis **research hypotheses are, research hypothesis are** because he or she has strong empirical or theoretical support. For doing this here she will use terms such as increased decrease or positive relationship or negative relationship and use the corresponding symbols. Shown here however if the researcher plans to state a non directional hypothesis **research hypotheses are, research hypothesis are** which is you is the hype of research hypothesis **research hypotheses are, research hypothesis are**. That usually used. The researcher will use terms or words such as effect difference a relationship and use these symbols listed here while we’re talking about stating hypotheses in words. I want to take a few minutes and talk about precise terminology. Although not all researchers would agree with this. Is they use terminology interchangeably. Some researchers suggest that the terms that are used in a hypothesis **research hypotheses are, research hypothesis are** and research questions imply specific research designs and statistical analyses. That are going to be used in the study therefore I would agree with these researchers and say that it is important that you use precise terminology. Let’s take a closer look at this. Let’s say that you state a null hypothesis **research hypotheses are, research hypothesis are** of there will be no significant relationship between variables. Now what design looks at relationship between variables a correlational design so if the word relationship is used in a hypothesis **research hypotheses are, research hypothesis are**.

This implies that you’re looking for a relationship and will likely use a correlational analysis or a correlational research design and the corresponding correlational analysis. Now let’s look at the word effect. Let’s say as a researcher you state that there will be no effect of a specific intervention on students achievement. That your null hypothesis **research hypotheses are, research hypothesis are** now. What research designs look for an effect. Well if you remember a true experimental and quasi-experimental design look or designs look for an effect between of an intervention. So chances are that you will. If you use the word effect you are going to do a quasi experimental design or a true experimental design and because these two designs look at the differences between groups or within groups. You’ll most likely do some type of group comparison analysis or difference analysis. Finally we have the word difference. Let’s say as a researcher you say that you are looking for a statistic or you are looking to see or test the null hypothesis **research hypotheses are, research hypothesis are** that there is no statistically significant difference between two groups and you’re looking to make sure that there’s no statistically significant difference in their academic achievement one group. Let’s say gets the intervention and one group doesn’t what designs look at differences. Well we just talked about – the quasi experimental design in the true experimental design right and so difference could imply a quasi experimental design or a true experimental design. It could also imply one other group. Comparison study it could imply a causal comparative design remember a causal comparative design. Can’t look for a cause and effect relationship but it can look for a possible cause and effect relationship so oftentimes researchers will use the word difference in their research questions and hypotheses when they’re using a causal comparative design. Now note here the word effect should never be used when talking about a causal comparative design.

Because you can never determine cause-and-effect relationship in a causal comparative design. You can only look at possible cause and effect so. Let’s consider one more example. Let’s consider that a researcher. You as a researcher posed the question. Is there a difference in university. Students perceived learning based on the type of course in which they participate either online or residential. What design first of all research design. Does this question imply. It could imply a true experimental design or quasi-experimental design it can also imply remember a causal comparative design and finally it also implies what type of analysis well it’s looking at the differences between two groups and that’s clearly stated on one type of or on one dependent variable so this implies some type of analysis of difference and in this case an independent t-test. Let’s look at one more question because it’s very important that you understand precise terminology and use precise terminology. What if you as a researcher posed the question or the hypothesis **research hypotheses are, research hypothesis are**. Let’s look at hypothesis **research hypotheses are, research hypothesis are** in this case a null hypothesis **research hypotheses are, research hypothesis are**. You’re going to test the null hypothesis **research hypotheses are, research hypothesis are** you’re going to test the null hypothesis **research hypotheses are, research hypothesis are** that there will be no statistically significant relationship between university students perceived learning scores and their sense of community scores. What first of all research design does this imply remember. You’re looking at relationship. It implies a correlational really research design and then what type of analysis does this null hypothesis **research hypotheses are, research hypothesis are** imply a correlational analysis. And since you have two scores maybe at ratio and interval level you can it implies the the researcher or you in this case plan to do a Pearson’s R okay. Now we know that the first step in hypothesis **research hypotheses are, research hypothesis are** testing is to state the null hypothesis **research hypotheses are, research hypothesis are** and in social science. We usually do this in words. And we’ve talked a little bit about how to do this in words but now we’re gonna take an even closer look at how to write an effective hypothesis **research hypotheses are, research hypothesis are**.

Both both a null hypothesis **research hypotheses are, research hypothesis are** as well as a research an alternative or alternative hypothesis **research hypotheses are, research hypothesis are** in words. So as you formulate your hypothesis **research hypotheses are, research hypothesis are** here are some criteria to ensure that you write an acceptable one and this is taken from. Bartos 1992 a well written hypothesis **research hypotheses are, research hypothesis are** states. The expected relationship between the two variables. It’s testable the hypothesis **research hypotheses are, research hypothesis are** is stated simply and concisely and finally the hypothesis **research hypotheses are, research hypothesis are** is founded in the problem statement and supported by the research. It aligns with the research question. We’re gonna take a look at how to do this. Let’s take a look at a hypothesis **research hypotheses are, research hypothesis are**. That was proposed by a beginning researcher. Now here you’ll see that the hypothesis **research hypotheses are, research hypothesis are** specifically the null hypothesis **research hypotheses are, research hypothesis are** was derived from the research question. And we’re going to assume in this case that it was derived from the empirical literature in the purpose statement so the researcher proposes two null hypotheses. We’re gonna take a look at the first one. The null hypothesis **research hypotheses are, research hypothesis are** states the media-rich technology education will not be effective for students. Now we’re gonna take a look at evaluating this using the criteria for an acceptable or well-written hypothesis **research hypotheses are, research hypothesis are**. Here we see the criteria for writing and acceptable hypothesis **research hypotheses are, research hypothesis are**. First of all does this hypothesis **research hypotheses are, research hypothesis are** state. The expected relationship between variables. Well we could say yeah probably the here we see. The researcher expects a relationship to exist between a media-rich technology education and effectiveness. However you’ve probably already thought what in the world do these. Two variables are these two concepts mean. The variables are not clearly defined. Thus what we would find is is it would be really really difficult to test them first of all. What is meant by effectiveness. Second of all what exactly is media-rich technology education. And finally. Who are these students. Are they primary students secondary students.

University students these terms in order to be testable need to be stated more simply more concisely more clearly. They need to be operationally defined in order for this hypothesis **research hypotheses are, research hypothesis are** to be testable remember. An operational definition is simply a clear concise definition of how a variable is going to be measured or observed in the research. Study this is different from a dictionary definition for example. Let’s look at the word effectiveness if we were to look up the word effectiveness in the dictionary we would find the following definition an adequate or if the following definition producing an intended or expected result to accomplish a purpose or adequately accomplishing a purpose. Now this dictionary definition is not precise enough for the purpose of research. So it’s too if we can say really it’s too general. What we need is a more exact indicator of effectiveness. This may be for example a score on some type of validated achievement test. Let’s say we were looking specifically here at math. We would want to look for a validated math achievement. Test or we may want to look. Let’s say effectiveness was being measured by some type of attitude toward math. We may want to look for a validated at an attitudinal survey such as the attitudes toward math inventory. That would be how we would specify or operationally define the word effectiveness. Next is the word word. Media-rich technology education this may be more precisely defined as the math – oh program and something then explained in the research study and let’s say that students can be more precisely defined. As undergrad graduate students enrolled in an online statistics course yeah in an online statistics course so understanding these operational definitions we can then revise them hypothesis **research hypotheses are, research hypothesis are** and make it more testable and also state it more simply and concisely and then finally the other two criteria is the hypothesis **research hypotheses are, research hypothesis are** grounded in the problem statement in the research and the research question we already established this and said yes it was so let’s take a look at a more or a better or well-written hypothesis **research hypotheses are, research hypothesis are**.

Now let’s take a look at restating. The null hypothesis **research hypotheses are, research hypothesis are** using what we just discussed that is. Let’s restate the null hypothesis **research hypotheses are, research hypothesis are** using the operational definitions in order to make our hypothesis **research hypotheses are, research hypothesis are**. More clear more concise and actually testable. There’s two ways you could restate it first of all you could state that the math tulo program will not statistically significantly effect online undergraduate students attitudes toward math as measured by the attitudes toward mathematics inventory. Another way to state it is is there will be no statistically significant difference in the mean final exam scores of online undergraduate students who participate in the math 2o program as opposed to the online undergraduate students who participate in a traditional statistics. Course now you can see that. There is a difference in the dependent variables here but you’ll also see another difference in the first one. We know that the researcher is examining the math 2o program in the second one we can clearly see the levels or groups of the independent variable that the researcher is examining so ultimately. I usually prefer to state null hypotheses in the second way so you can clearly see not only the population being studied and the dependent variable being studied but also the groups or the levels of the independent variable being studied. Now you’ll note in the two well written or acceptable hypotheses that we just looked at we use the term significant or statistically significant or not statistically significant and no significance. Remember when. You’re testing null hypotheses or you’re testing hypotheses you’re examining to see not only if there is a difference but a statistically significant one. You’re observing whether the difference was due to chance or the intervention therefore when you’re writing your know hypotheses as well as your research and alternative hypotheses it’s important to use the term no statistically significant or statistically significant or no significant or not significant or significant rather than just using the term difference.

Because you were looking for a statistically significant difference now. We’ve talked a lot about writing a well-written hypothesis **research hypotheses are, research hypothesis are**. We talked about using precise terminology that implies the type of design and analysis that we have planned for our research study we talked about using operational definitions to ensure that our hypotheses are stated. Clearly concisely and are testable now believe it or not. We are not done talking about writing. Hypotheses remember we’re still on step one of hypothesis **research hypotheses are, research hypothesis are** testing stating the null and research hypothesis **research hypotheses are, research hypothesis are**. There’s one more thing we need to talk about before we move on and that is the rules about the number of hypotheses needed for each research question you now. Every research question needs to have at least one corresponding null hypothesis **research hypotheses are, research hypothesis are** and one corresponding research hypothesis **research hypotheses are, research hypothesis are** however because of the choice of analysis for a research study sometimes more than one hypothesis **research hypotheses are, research hypothesis are** is needed more than one research hypothesis **research hypotheses are, research hypothesis are** and more than one null hypothesis **research hypotheses are, research hypothesis are** is needed for one question. The number of hypotheses is dependent upon the analysis chosen but the number of hypotheses can be determined based upon the number of variables under investigation in the study. Let’s take a look at the first role and this one’s pretty simple. If you’re conducting research in which differences between groups are examined and you have one independent variable and one dependent variable then one research hike or research question is needed with one corresponding null hypothesis **research hypotheses are, research hypothesis are** and one corresponding research hypothesis **research hypotheses are, research hypothesis are**. Because only one main effect is studied in a correlational. Study if you have two variables of interest you’re just looking at the relationship between two variables. Then you only need to write one null hypothesis **research hypotheses are, research hypothesis are** and one research hypothesis **research hypotheses are, research hypothesis are** to correspond with your research question.

Let’s take a look at this here. We see a researcher who’s interested in understanding if there is a difference in students sense of community scores based on the type of course in which they participate since here the intro. The researchers interested in differences in one dependent variable which is sense of community based on one independent variable which is type of course and this independent variable may have two groups online and residential or it may have three groups online residential and. Let’s say hybrid. There’s one independent variable and one dependent variable in this research question then. The researcher only needs to pose one corresponding research or alternative hypotheses hypothesis **research hypotheses are, research hypothesis are** and one null hypothesis **research hypotheses are, research hypothesis are**. Here we see an example of the null hypothesis **research hypotheses are, research hypothesis are** the null hypothesis **research hypotheses are, research hypothesis are** states. There is no statistically significant difference in sense of community among undergraduate students based on type of course this hypothesis **research hypotheses are, research hypothesis are** would examine one main effect. Now let’s take a look at a correlational example. Here we see instead of being interested in the difference between groups. The researcher is interested in the relationship between variables and plans to conduct a correlational study. Specifically the researcher is interested in if there is a statistically significant relationship between undergraduate students perceived learning scores and their sense of community since the research question has two variables of interest perceived learning and sense of community. Only one null hypothesis **research hypotheses are, research hypothesis are** is needed in one research. Hypotheses and here you see an example null hypothesis **research hypotheses are, research hypothesis are**. There is no statistically significant relationship between undergraduate students perceived learning and their sense of community. Now let’s take a look at the second rule. The second rule is this if you have two independent variables in one dependent variable you test three null hypotheses you test two main effect null hypotheses one for each of your independent variables and an interaction interaction hypothesis **research hypotheses are, research hypothesis are**.

Which is referred to as a first-order interaction. The interaction between two variables. So if your research question has two independent and one dependent variables you state three null hypotheses and you state three alternative hypotheses. Let’s take a look at this role here. In this example. We see that the researcher is still interested in examining student sense of community. And he’s still interested in examining student sense of community based on the type of course in which they’re enrolled however he’s also interested in another independent variable gender so he’s interested in how sense of community differs based on whether or what type of course the students taking as well as their gender so he poses the question will there be a statistically significant difference in student sense of community scores based on the type. Of course they take and their gender so here since the researcher has one dependent variable sense of community and two independent variables type of course and gender then. The researcher needs to pose three separate null hypotheses. He’ll also have to say three separate alternative hypotheses. Let’s take a look at these null hypotheses and their stated here for you first of all the researchers going to have to pose a main effect hypothesis **research hypotheses are, research hypothesis are** for the first factor or independent variable so the first one states this there is no statistically significant difference and sense of community among students based on their type of course they take whether it be online or residential. The researcher also needs to state a main effect hypothesis **research hypotheses are, research hypothesis are** for the second independent variable or factor. This null hypothesis **research hypotheses are, research hypothesis are** would state. There’s no statistically significant difference in sense of community among students based on their gender whether they’re male or female the researcher would also need to state and interaction hypothesis **research hypotheses are, research hypothesis are** which considers both independent variables and this would be stated like this.

There is no statistically significant difference in sense of community among students based on the type. Of course they take while their gender remains constant or and based on gender so here you see an example of how to write hypotheses when you have one dependent variable and two independent variables. Now what if in our research scenario the researcher wants to also look at ethnicity and add a third independent variable. Well if you have three independent variables and one dependent variable you. Test seven null hypotheses now. Why seven null hypotheses. Well this time you would have three main effect null hypotheses in this case one for gender one for ethnicity and one for type of course three first-order interaction. Hypotheses so one hour so hypotheses that look at the interaction between each of those variables between gender and ethnicity gender and type of online course and then ethnicity and type of online course and then finally a second-order interaction which involves all three independent variables. So if you have three independent variables in one dependent variable. You would need to plan to test seven null hypotheses. So you’d state seven null hypotheses in step seven research or alternative hypotheses so each time you increase the number of independent variables you increase your number of null and research. Hypotheses you increase both the main effect hypotheses as well as the interaction hypotheses so we’ve looked at independent variables. Now let’s take a look at increasing the number of dependent variables if you plan to examine two or more related dependent variables then you need to state a hypothesis **research hypotheses are, research hypothesis are** a null hypothesis **research hypotheses are, research hypothesis are** for the linear combination of the variables as well as each variable separately. And then you would also need to state research hypotheses for the linear combination of the variables as well as for each variable separately now note. I said here correlated or related dependent variables if dependent variables are significant significant significantly related.

Then they need to be analyzed together in order to control for possible air if however they’re not related then you would just consider them separately and you’d have let’s say one independent variable and one dependent variable so you go back to one hypothesis **research hypotheses are, research hypothesis are** now. The same logic holds true for correlational studies if you have two or more variables of interest or two or more predictors. You need to state hypotheses for the linear combination of the variables as well as each variable separately. Let’s go ahead and nail take a look at this. Let’s say that the researcher in the example that we’ve been considering once to not only consider a sense of community but he wants to look at the subscales for community which are learning community and connectedness and through a literature review. He determines that these two variables are related so he knows that they need to be considered together in his research question and ultimately his hypotheses and then analysis so he poses the question will there be a statistically significant difference in learning community and connectedness among undergraduate students based on the type of course in which they’re enrolled so here we see that he has the independent variable again of type of course and it has two levels or two groups and he has two dependent variables like we said learning community and connectedness so based on the rule that we just went over. We know that he needs to have three null hypotheses and three researcher alternative hypotheses because he needs to consider remember the linear combination of the variables as well as he needs to consider in his hypotheses each variable separately and since he has two dependent variables that are related. He needs one hypothesis **research hypotheses are, research hypothesis are** for both of them and one for each individually so these then become his null hypotheses first of all. He looks at the linear combination in.

He says there’s there will be no statistically significant difference in the linear combination of learning community and connectedness among undergraduate students based on their type of course this considers both of the variables. Then he looks at each individually. There will be no statistically significant difference in learning community among undergraduate students based on their type of course and then there will be no statistically significant difference in connectedness among undergraduate students based on their type. Of course now what would happen if let’s say sense of community had three sub skills and not just two. How many null. Hypotheses and alternative hypotheses. Would the researcher need in this case. If you’re thinking four of each you’re correct. He needs four for null hypotheses and for research hypotheses because he’ll need one hypothesis **research hypotheses are, research hypothesis are** to consider the linear combination of the variables. And then one for each of the three individually. Now let’s take a look at this rule or you looking at a correlational study here. We see the researcher wants to know if students sense of community scores can predict perceived learning and course points so the researcher is interested in examining the relationship between one criterion variable sense of community and two predictor variables perceived learning and course points therefore based on the rule that we discussed the researcher needs three alternative hypotheses and three null hypotheses and the three null hypotheses are listed here first. We see a null hypotheses hypothesis **research hypotheses are, research hypothesis are** for the combination of the two predictor variables its States online undergraduate student. Sense of community does not statistically significantly predict their perceived learning and course points then we see a hypothesis **research hypotheses are, research hypothesis are** for each of the predictors individually first for perceived learning and then for course points so this provides an example. We’re going to look at one more rule before we move on to the next step in hypothesis **research hypotheses are, research hypothesis are**.

Testing the final rule for writing a hypothesis **research hypotheses are, research hypothesis are** or hypotheses. Is this when you have a control. Variable this needs to be reflected in the hypothesis **research hypotheses are, research hypothesis are** or hypotheses for example. Let’s say that the researcher is interested in knowing if there’s a difference in sense of community among students based on the type of course in which they’re enrolled either online or residential. But he also wants to control for one other variable and that’s gender here you can see an example null hypothesis **research hypotheses are, research hypothesis are** that considers the control variable or the covariant. It states this. There will be no statistically significant difference in sense of community among students based on type of course online or residential while controlling for students gender that while controlling for students gender reflects the control variable. So now we’ve talked about how to write good null hypotheses and good alternative or research hypotheses. We talked about writing imprecise language. Using correct terminology we talked about using operational definitions to in order to make a hypothesis **research hypotheses are, research hypothesis are** testable and then finally we talked about the rules for the number of null and research hypotheses that are needed for each research question. We’re now going to move on to step two of hypothesis **research hypotheses are, research hypothesis are** testing as we’ve been talking about hypotheses and hypothesis **research hypotheses are, research hypothesis are** testing for quite a while. Now we’re going to look at steps two through four of hypothesis **research hypotheses are, research hypothesis are** testing as well as type 1 and type 2 error in the next tutorial. So this concludes this tutorial for now you.

**Where to find great research papers?**

Various great research journals such as Global Research Letters are a great option and way to help you look up impactful research papers with a great format. Here, you will find a number of various research papers that are provided and made available to you in the journal, which will help you write your own paper.

You can very easily find papers on a variety of topics at Global Research Letters, which will help you with your own research work and understanding of writing and publishing research papers properly. With access to so many amazing research papers, you can practice and learn the process of writing research papers and their importance.