Samenvatting
Summary all lectures MRM2 + notes : Grade 8.5
- Instelling
- Universiteit Van Amsterdam (UvA)
Summary all MRM 2 lectures notes and a lot of examples to understand the material better. Grade 8.5
[Meer zien]Voorbeeld 4 van de 51 pagina's
Voorbeeld 4 van de 51 pagina's
In winkelwagengesponsord bericht van onze partner
3 beoordelingen
Door: MelissaMajoor • 5 jaar geleden
Door: hrconsultingasunta • 6 jaar geleden
Door: inalom • 6 jaar geleden
Voorbeeld van de inhoud
Lecture 1: analysis of varianceRecap MRM1
• p-value = the probability that your 0-hypothesis is true, given your data. With a p-value lower
than 5%, we are going to reject the 0-hyp.
• T-test has it’s assumptions: N>30 and data should be normally distributed. Only if those
conditions are met you can do a valid t-test.
• Central limit theorem when sample size goes up data gets approximately normally
distributed.
• A shapiro-Wilk or a Kolmogorov-Simornov test can help to find out whether we can assume
that data follow a normal distribution you want the p-value to be bigger than 0.5 of 0.10
(depending on the level of significance you choose).
Week 1: ANOVA
Conceptual models are visual representations of relations between theoretical constructs (and
variables) of interest. You hypothesise a relation between predictive and outcome/dependent
variable and this can be shown in a conceptual model.
OV=outcome variable dependent
PV= predictor variable independent variable
OV and PVs can be categorical and quantitative.
Lecture example: what factors drive online mobile spending?
H0= people who own a mobile phone don’t spend
more via mobile spending.
H1 = People who own a mobile phone spend more
via mobile spending.
This is the most simple version of drawing a conceptual model, but there are often moderators and
mediators involved (in the effect on the OV).
Moderation
Maybe there is an effect between having a mobile phone and
online spending, but is it also caused by the opportunity to
pay via the mobile phone. “Option to pay via mobile” is a
moderating variable one variable moderates the
relationship between two other variables. In case of
moderation you want to understand the when question.
1
,Mediation
What if the proposed relationship “goes through” another
variable? This leads to H3 = The positive effect of mobile
ownership on online mobile spending is mediated by mobile
browsing. “Mobile browsing” is a mediating variable one
variable mediates the relationship between two other
variables.
Mediation is different from moderation. Mediation explains an underlying meganism. Mediation is
more the why question.
In research papers you will find multiple mediators/moderators in a model, which can make
conceptual models look more difficult than they actually are. In general there are a few rules:
• Boxes always represent variables (so not a level of a variable colour is variable and colours
are levels.
• Arrows represent relationships between variables from PV to OV.
• Dotted line is the mediation effect. You first test the direct effect and replace this by the
mediator.
• You can see that there are
only 2 independent variables in this
conceptual model: colour and
sensation seeking
The theory behind an Anova Test
Anova means analysis of variance. Variance spread of the observations to the mean. So it’s the
average of the squared differences from the mean, and it is squared because you want to get rid of
the negative values. If we have two data points with scores 2 and 3, the mean score = 2.5, hence the
variance =
If you run an Anova, you want to examine how much of the variance in the data can be explained by
the independent variable. We use (between-subjects) ANOVA when:
- OV = Quantitative
- PV = Categorical with > 2 groups
- Between-subjects designs groups in your sample should be independent on each other.
120 participant; only participate in 1-group at the same time.
- Variance is homogenous across groups
- Residuals are normally distributed
2
,Example research question: Is there a relation between shopping platform and customer
satisfaction? You have 3 channels. PV (shopping platform is categorical) OV=quantitative (1-5 scale).
Conditions: PV=categorical with more than 2 levels and quantitative outcome variable, so you can do
an Anova
• H0 = There is no difference in outcome variable scores between different levels of the
independent variable. μ1 = μ2 = … = μi
• H1 = There is a difference in outcome variable scores between at least two levels of the
independent variable.
We have 10 observations on customer satisfaction scores (OV). The grand overall mean is 32.3 and
the total Sum of Squares =1192.10
Step 1: Calculate the total distance to
the mean (sum of squares) square
it to get rid of negative variance. So
the total variance in the sample =
1192.10. But how much of this
variance can you explain by your
predictive variable?
Step 2: Now you are going to regroup
them based on the outcome (mean).
All these groups have a group mean as
well. So the distance from group
mean to the total mean is basically
the part of the variance that you’re
explaining by the model.
3
, Step 3: you see that not all the
observations are on the group
mean as well (not on one line).
This is not incorporated in your
model. So the distance from the
variables to the group mean is
the variance that we are not
explaining in our model. Than
you sum up all the distance from
group means and square that.
Step 4: We have now divided the variance in our data in a part that can be explained by our model
(between SS) and a residual part (within SS). You would expect that the sum of the squares (total
variance)= the sum of squares that you are explaining and the one that you are not explaining. But
you are only interested in the part of the variance in the data that you can explain, because only if
this is significant, you can say something about the causality of the model.
Step 5: We can now calculate how much of the total variance in our data is explained by our model.
This ratio is called R2. You can do that by the formula (the part that you explain/by total variance).
Step 6: However, the Sum of Squares is affected by the number of observations. We therefore need
to calculate the Mean Squares (for both the model and the residuals).You divide this by the degrees
of freedom. So total variance, divided by number of groups that you have in the independent
variable.
Step 7: The F-ratio expresses the proportion of explained variance relative to unexplained variance.
You calculate the ratio of explained variance, by not explained variance.
F you want it to be as high as possible. F=high means that a lot of the variance can be
explained by the model. You can test if an F-ratio is significant SPSS gives this. There is no rule of
thump; F-ratio should be above.. It depends on the degree of freedom. You will get a F-level of
significance.
4
Voordelen van het kopen van samenvattingen bij Stuvia op een rij:
Verzekerd van kwaliteit door reviews
Stuvia-klanten hebben meer dan 700.000 samenvattingen beoordeeld. Zo weet je zeker dat je de beste documenten koopt!
Snel en makkelijk kopen
Je betaalt supersnel en eenmalig met iDeal, creditcard of Stuvia-tegoed voor de samenvatting. Zonder lidmaatschap.
Focus op de essentie
Samenvattingen worden geschreven voor en door anderen. Daarom zijn de samenvattingen altijd betrouwbaar en actueel. Zo kom je snel tot de kern!
Veelgestelde vragen
Wat krijg ik als ik dit document koop?
Je krijgt een PDF, die direct beschikbaar is na je aankoop. Het gekochte document is altijd, overal en oneindig toegankelijk via je profiel.
Tevredenheidsgarantie: hoe werkt dat?
Onze tevredenheidsgarantie zorgt ervoor dat je altijd een studiedocument vindt dat goed bij je past. Je vult een formulier in en onze klantenservice regelt de rest.
Van wie koop ik deze samenvatting?
Stuvia is een marktplaats, je koop dit document dus niet van ons, maar van verkoper ambervandeelen. Stuvia faciliteert de betaling aan de verkoper.
Zit ik meteen vast aan een abonnement?
Nee, je koopt alleen deze samenvatting voor €7,49. Je zit daarna nergens aan vast.
Is Stuvia te vertrouwen?
4,6 sterren op Google & Trustpilot (+1000 reviews)
Afgelopen 30 dagen zijn er 66579 samenvattingen verkocht
Opgericht in 2010, al 14 jaar dé plek om samenvattingen te kopen