Reliability in SPSS - Coefficient alpha (Part 1)

Get the SPSS help you need fast for an amazingly low price: tinyurl.com/oo3...
Act now - offer ends soon.
Learn how to calculate and interpret a reliability coefficient in SPSS using Coefficient (Cronbach's) alpha (Part 1)
Coefficient alpha
Cronbach's alpha
internal consistency reliability
How to Calculate and Interpret Coefficient Alpha (Cronbach's alpha) in SPSS is shown in this video
Cronbach's Alpha
Coefficient Alpha
Internal Consistency
Reliability
In this video, we'll take a look at Coefficient alpha, and how to run that in SPSS. Coefficient alpha is one way to estimate the reliability of an instrument (internal consistency reliability). In this example, we'll look at the satisfaction with life scale (Diner et al, 1985), which consists of 5 items on a 7-point Likert scale. You are probably familiar with Likert scales and have taken them before. With 5 questions on a 1 to 7 scale, the lowest possible total score is 5 to a high possible score of 35 (answering 7 on every question). Any total score outside of this range (5 to 35) would indicate that an error had been made.
In SPSS, there are 5 different variables with responses for 80 total people. To run the analysis, we want to select Analyze, then Scale, followed by Reliability Analysis. Move all 5 variables (items) to the item box. We'll add the label "Satisfaction with Life Scale." Next, click Statistics and select all the values under Descriptives for section of the table. Click Continue, then OK.
Our second table of output, Reliability Statistics, contains the value we are most interested in, Cronbach's Alpha. Our reliability estimate is .88, which is very good. (See part 2 for further discussion of the output.)
KZbin channel: / statisticsinstructor
Udemy inferential course: www.udemy.com/...

[@bendebulejoseph5405]

its the best way of achieving all statistical analysis

[@ummaru8776]

Highly inspired. The real Master who can master others. Your explanation and application is very simple and quick to grasp. Everything is clearly understandable as you proceed step by step. Stay blessed ever. Upload more videos Sir. Prayerful wishes and Love from India.

[@dr.ratanbisui9658]

Super VDO for learning. Thanks

[@purvitripathi8843]

So helpful .... Thanks ....i am a beginner so very helpful it is..

[@DavidFeral]

Very helpful. Thank you.

[@aylafil9042]

Thank you for your video. How can I calculate ceiling and floor effect? Nobody has explained it yet. It would be more useful to test the reliability of scales.

[@anitacarrier9386]

I have a 5 point likert scale, does this make a difference? Also i dont see where you put in your minimum and maximum answer scores so how does spss work out if it's reliable?

[@ChillMall]

Helpful.. Keep it up teacher

[@fernandoalberto1635]

Great video, thanks.

[@sumayyahradzuan8383]

What to do if I got Excluded(there is amount) in the case processing summary table? help me

[@dharrsheena]

Thank you statisticsinstructor for tis video. I have a question. My questionnaire has 6 questions. Num 1 is a Yes or No question in which the respondent will have to skip Num 2, Num 3 and Num 4 if they answered No for Num 1. I cannot calculate Cronbach Alpha as they are missing values. Should I give them a value? (for exmaple 0 for skip, 1 for yes, 2 for no). So tat it wont be considered as missing values

[@trxyej283]

hey I just watched your video and I hope you can help me, I have several questions in my survey, we've done a pilot study already but the downside to it is that most of my questions don't have a Likert scale and mostly the videos I saw uses a Likert scale mostly my questions are open-ended and dichotomous like yes/no questions will I still able to test the reliability of my instrument? please help TT thank you so much!

[@adilahhassan3947]

Hi Quantitative Specialists Thank you for explaining this.But i have question.I have a 55 items questionnaire separated into 6 sub-sections. 5 of the subsections are using 4-point likert scale "Strongly Agree","Agree","Disargee","Strongly Disagree". Another one subsection is using "Never","Sometimes","Often","All the time". So how do i insert the data into SPSS and compute the reliability test to get the Cronbach's alpa?Should i do it separately?Please help me.Thank you.

[@QuantitativeSpecialists]

***** Adela, You should check the scale author(s) for instructions. They should indicate how to score it - Go to their original artcile in which the scale was published and you should see it there.

[@homaansari9221]

Can u plz tell me what would be the no of respondents for checking the reliability of the instrument??

[@snag57]

Hi. If you want to score True/False items for the KR-20 in SPSS how do you do it? Do you do Correct=1, Incorrect=0 or True=1, False=0?

[@janitazulfiqar3273]

How to check reliability of subscale

[@anaismendieta221]

Thank you you are the best. This video helps a lot

[@QuantitativeSpecialists]

+anais mendieta Thanks - glad to hear you found it helpful!

[@missmoody7409]

My all variables reliability alpha cronpach's is. 935. So how can i write in interpretation kindly guide me.

[@sabhyajuneja]

can i run reliability test having 200 variables and data of 10 respondents?

[@tiffanyyu7484]

Hi sir, would you plz tell methe factors that contribute to coefficient alpha?

[@QuantitativeSpecialists]

Tiffany yu Hi TIffany, The easiest way to think about it is that alpha is influecned by two things: (1) the correlation among the items (biggest influence) and (2) the number of items, k. As the number of items increase (for a fixed correlation), alpha goes up. This tends to have the most influence when k is realtively small (8 or less). I hope this helps. QS

[@TheAmandeep2009]

very good

[@madees6602]

Hi statisticinstructor, I had a hard time understanding the reliability part, isn't it we should test the reliability before we use the instrument? thus, how do we test the reliability without the data collections yet?

[@QuantitativeSpecialists]

Hi Little Suang, We need to use the instrument in order to get the data to assess reliability. What happens is if the reliability isn’t good, then the instrument should be modified or abandoned. Most of the time though we use published instruments that already have shown to have adequate or good reliability. But if we are making our own instrument or using an unpublished one, then poor reliability should lead to modifying the instrument (and getting data on a new sample) or using a different one. I hope this helps.

[@madees6602]

thank you so much! =)

[@meeramenonentrepreneurclasses]

Quantitative Specialists Hi do you have examples using factor analysis and multiple regression for data analysis

[@QuantitativeSpecialists]

Meera Menon Hi Meera, You can find factor analysis here: kzbin.info/www/bejne/l5vZoXx3eKismpY and multiple regression here: kzbin.info/www/bejne/qpmykIqcmchphZY Let me know if they help you out with your data analysis...

[@meeramenonentrepreneurclasses]

It was useful. Thank you.any examples with 2 dependent variables and 4 independent predictors

[@JH-nr9pw]

thank you SO MUCH YOU ARE THE BEST

[@QuantitativeSpecialists]

+Rose Chiba My pleasure, Rose! Please spread the word!

[@fieqafie7676]

Hi +Quantitative Specialist. Thank you for explaining this. I have a question. I have a questionnaire consist of 6 question and have open ended questionnaire and also a question that are using scale such as "never" , "seldom' "often" , "very often" and other scale too. can i use this test to test the reliability of my questionnaire? Please help me. Thank you

[@QuantitativeSpecialists]

fieqa fie Hi fieqa fie, My pleasure. We have videos on assessing the reliabilty of a questionnaire here: @2 If you take a look at these I think it should help out. QS

[@Deannn__]

Hi, is it acceptable if my cronbach alpha is 0.99? or do i need to adjust?

[@sanabatool6061]

hey i think you have to adjust because its too much high reliability of any scale. which scale you were using? and what is the reliability of the original scale?

[@momoberry_]

Hi sir, i'd like to know if .740 is still considered reliable for my data? what is the average range?

[@QuantitativeSpecialists]

Hi Joanne, .76 is fine. In terms of the range, most people consider .70 and above as adequate, with some feeling .60 and above as fair. My preference is for .70 and higher. But either way, you're in good shape! If you're looking for a reference, Nunnally (1978) is a good one.

[@momoberry_]

Thank you so much!

[@hanialbarni3925]

if my cronbach's alpha is 0.4 what does mean 0.5 , 0.6 , 0.7 , 0.8 ?

[@QuantitativeSpecialists]

Hani Albarni .4 would be considered low, meaning that the items are not very internally consistent. .5 is higher (more internally consistent), but still low. .7 is better.

[@lmb1234]

Hey your videos are very helpful I'm currently working through them. I am currently doing my dissertation and I have used a SERVQUAL survey for my questionaire. This basically shows me consumer expectations and perceptions of service quality, and rates it 1-5. Somehow when doing my coefficient alpha I've received a score of .910 and I think I've done something wrong. I have 100 participants and the first section of my survey is demographic details and I havent included those questions in this would that be right or wrong? I would really appreciate your help, Thank you Quantitative Specialists

[@QuantitativeSpecialists]

laurenbuchanan2 Hi Lauren, Without being able to see your survey or data, it sounds like you are probably performing things correctly. However, to run coefficient alpha you are assuming a unidimensional scale, so you want to make sure your survey is measuring the same concept (construct) throughout. If this is the case, .91 is quite good! QS

[@summerrosesoapbymimie3183]

cannot see. its sensord

[@QuantitativeSpecialists]

It's showing up on my end. Perhaps trying it again might work.

[@kdsboysl4744]

Molorahaknaaa