Deutsches R-Forum

Hallo zusammen,

ich möchte mit R Studio eine Metaanalyse berechnen und habe dazu in einem ersten Schritt die Excel-Tabelle eingelesen.
Nun möchte ich aus den deskriptiven Daten (N, M, SD) Hedges g berechnen und wende dazu die escalc-Funktion an.
Da dabei superviele Warnungen aufgetaucht sind, habe ich mich zunächst auf eine Studie (Effect_Size_No == "17") beschränkt, die meiner Meinung alle Daten beinhaltet, die ich brauche.

> data <- escalc(measure="SMD", subset = Effect_Size_No == "17", n1i = N_Women, n2i = N_Men, m1i = Mean_Female, m2i = Mean_Men, sd1i = SD_Female, sd2i = SD_Men, data=data2)

Allerdings kommen dabei folgende Warnungen:


Warning messages:
1: In Ops.factor(sd1i, 2) : ‘^’ not meaningful for factors
2: In Ops.factor(sd2i, 2) : ‘^’ not meaningful for factors
3: In Ops.factor(m1i, m2i) : ‘-’ not meaningful for factors

Hat jemand von euch damit Erfahrung? Ich weiß absolut nicht, was das zu bedeuten hat. Alle Zellen, die ich angegeben haben, beinhalten reine Zahlenwerte, mit . getrennt (z.B. 1.23).

Vielen Dank im Voraus!

Liebe Grüße

Hallo Antoniia13,

willkommen im Forum!
Aus welchem Paket ist die Funktion escalc()
Kannst Du bitte den Output von in Deine nächste Nachricht kopieren?
Vielleicht ist beim Einlesen der Daten irgendwas schief gelaufen ...

Gruß, Jörg

Die ist aus dem Paket metafor.

Hier der Output mit str(data2) - ich hoffe, du kannst etwas damit anfangen.

Danke und viele Grüße!

> str(data2)
'data.frame': 51 obs. of 56 variables:
$ Whose.study : Factor w/ 3 levels "","A","F": 3 3 3 3 3 3 3 3 3 3 ...
$ Article_No : int 2 2 2 3 3 4 4 5 6 7 ...
$ Sample_No : int 2 2 2 3 3 4 5 6 7 8 ...
$ Effect_Size_No : int 2 3 4 5 6 7 8 9 10 11 ...
$ Reference... : Factor w/ 25 levels "","Banai, M., Stefanidis, A., Shetach, A., & ™zbek, M. F. (2014). \nAttitudes toward ethically questionable negotiation tactics: \"| __truncated__,..: 2 2 2 3 3 4 4 5 6 7 ...
$ Study.No. : Factor w/ 7 levels "","/","1","2",..: 1 1 1 1 1 3 4 1 1 3 ...
$ Year : int 2014 2014 2014 2016 2016 2010 2010 2011 2013 2009 ...
$ Continent.. : Factor w/ 10 levels "","America","America ",..: 6 6 6 5 5 2 2 9 2 2 ...
$ Country : Factor w/ 13 levels "","Brazil","Canada",..: 8 8 8 4 4 12 12 11 2 12 ...
$ MAS_Index : Factor w/ 11 levels "","42","45","47",..: 5 5 5 11 11 9 9 3 6 9 ...
$ Type_Of_Source.. : int 1 1 1 1 1 1 1 1 1 1 ...
$ Published.. : int 1 1 1 1 1 1 1 1 1 1 ...
$ N_Total : int 615 615 615 173 173 379 172 361 230 248 ...
$ Percentage_Women : Factor w/ 41 levels ""," 61.0 ","0.00%",..: 3 3 3 2 39 40 5 11 18 20 ...
$ N_WomenXXX : num 0 0 0 106 106 ...
$ N_Women : num 0 0 0 106 106 239 39 120 101 114 ...
$ N_Men : num 615 615 615 67 67 140 133 241 129 134 ...
$ Average_Age : Factor w/ 28 levels "","/","19.00",..: 2 2 2 2 2 3 13 25 26 9 ...
$ Kind_Of_Sample : Factor w/ 8 levels "","MBA","MBA\nWorking ",..: 8 8 8 7 7 7 2 8 8 7 ...
$ Behavior_Vs_Attitude : Factor w/ 3 levels "","attitude",..: 2 2 2 2 2 2 2 2 3 2 ...
$ Kind_Of_Negotiation : Factor w/ 5 levels "","face_to_face",..: 3 3 3 3 3 3 3 3 5 3 ...
$ Assessment_DV_UB... : Factor w/ 27 levels "","1","1 (two scales)",..: 8 8 8 27 27 27 27 21 16 3 ...
$ Variable_Characteristic : Factor w/ 3 levels "","binary","continuous": 3 3 3 3 3 3 3 3 1 3 ...
$ Context_Of_Neg : Factor w/ 11 levels "","/","/ ","Car_Selling",..: 2 2 2 2 2 2 2 2 8 2 ...
$ Score_Type...average..subscore : Factor w/ 4 levels "","/","average",..: 4 4 4 3 3 3 3 1 1 4 ...
$ Reliability_Of_Scale : Factor w/ 30 levels ""," .89",".57-.73",..: 15 6 7 23 24 27 21 18 30 8 ...
$ Kind_of_UB... : Factor w/ 5 levels "","/","Active",..: 2 2 2 2 2 2 2 2 2 2 ...
$ UB_Specified : Factor w/ 15 levels "","(A)","(A), False Pr.",..: 13 11 12 7 15 6 6 6 5 10 ...
$ Mean_Female : Factor w/ 33 levels "","-0.02","/",..: 28 26 16 25 25 22 21 1 1 1 ...
$ SD_Female : Factor w/ 29 levels ""," 1.25",".63",..: 16 20 2 10 11 4 3 1 1 1 ...
$ Mean_Men : Factor w/ 33 levels "","-0.05","/",..: 29 22 10 27 28 25 26 1 1 1 ...
$ SD_Men : Factor w/ 30 levels "",".75",".86",..: 17 20 22 9 8 3 2 1 1 1 ...
$ Mean_Gender : Factor w/ 12 levels "",".20",".27",..: 10 10 10 10 10 10 10 1 1 4 ...
$ SD_Gender : Factor w/ 9 levels "",".40",".43",..: 7 7 7 7 7 7 7 1 1 5 ...
$ Mean_UB : Factor w/ 16 levels "","-.36",".22",..: 4 4 4 4 4 4 4 1 1 12 ...
$ SD_UB : Factor w/ 15 levels "",".42",".95",..: 4 4 4 4 4 4 4 1 1 13 ...
$ Frequency_Use_UB : Factor w/ 7 levels "","/","M: 13/51 (25%)\nF: 5/45 (11%)",..: 2 2 2 2 2 2 2 1 1 2 ...
$ Direction_Gender_Coding : Factor w/ 5 levels "","/","female",..: 2 2 2 2 2 3 3 1 1 3 ...
$ Direction_Unethical_Behavior : int 1 1 1 1 1 1 1 1 1 1 ...
$ Correlation_Gender_UB : Factor w/ 16 levels "","-.02\n","-.03",..: 15 15 15 15 15 15 15 1 1 10 ...
$ Other_Values : Factor w/ 10 levels "","/","Chi-Quadrat (1, N=96)\n = 3.25, p=.07",..: 2 2 2 2 2 2 2 1 1 1 ...
$ Effect_Size : logi NA NA NA NA NA NA ...
$ Advocacy : Factor w/ 5 levels "","/","no","yes",..: 2 2 2 2 2 2 2 2 3 2 ...
$ Financial_Incentive : Factor w/ 4 levels "","/","no","yes": 2 2 2 2 2 2 2 2 3 2 ...
$ Type_Of_Negotiation : Factor w/ 4 levels "","/","distributive",..: 2 2 2 2 2 2 2 2 4 2 ...
$ Experience : Factor w/ 4 levels "","/","no","yes": 4 4 4 3 3 3 4 4 4 3 ...
$ Opponent_Sex : Factor w/ 3 levels "","/","no": 2 2 2 2 2 2 2 2 3 2 ...
$ Instruction_To_Maximize : Factor w/ 4 levels "","/","no","yes": 2 2 2 2 2 2 2 2 3 2 ...
$ Acquaintance : Factor w/ 3 levels "","/","no": 2 2 2 2 2 2 2 2 3 2 ...
$ Expectation_Future_Interaction : Factor w/ 4 levels "","/","no","yes": 2 2 2 2 2 2 2 2 4 2 ...
$ Gender_First_Author : Factor w/ 3 levels "","female","male": 3 3 3 2 2 2 2 3 2 2 ...
$ Explicit_Hypothesis_Gender_Differences: Factor w/ 3 levels "","no","yes": 2 2 2 3 3 2 2 2 2 2 ...
$ Steterotype_Activation : Factor w/ 3 levels "","/","no": 2 2 2 2 2 2 2 2 3 2 ...
$ Threat_Gender_Status : Factor w/ 3 levels "","/","no": 2 2 2 2 2 2 2 2 3 2 ...
$ Power_Differences : Factor w/ 4 levels "","/","no","yes": 2 2 2 2 2 2 2 2 3 2 ...
$ Mentioning_.Possibilty_UB : Factor w/ 4 levels "","/","no","yes": 2 2 2 2 2 2 2 2 3 2 ...

Antoniia13 hat geschrieben: ↑Mi Aug 29, 2018 8:38 am Die ist aus dem Paket metafor.

Hier der Output mit str(data2) - ich hoffe, du kannst etwas damit anfangen.

ja, sicher.
Da gibt es einige Variablen, die Du sicher nicht als Faktor haben wolltest:

> str(data2)
'data.frame': 51 obs. of 56 variables:
$ MAS_Index : Factor w/ 11 levels "","42","45","47",..: 5 5 5 11 11 9 9 3 6 9 ...
$ Percentage_Women : Factor w/ 41 levels ""," 61.0 ","0.00%",..: 3 3 3 2 39 40 5 11 18 20 ...
$ Average_Age : Factor w/ 28 levels "","/","19.00",..: 2 2 2 2 2 3 13 25 26 9 ...
$ Assessment_DV_UB... : Factor w/ 27 levels "","1","1 (two scales)",..: 8 8 8 27 27 27 27 21 16 3 ...
$ Reliability_Of_Scale : Factor w/ 30 levels ""," .89",".57-.73",..: 15 6 7 23 24 27 21 18 30 8 ...
$ Mean_Female : Factor w/ 33 levels "","-0.02","/",..: 28 26 16 25 25 22 21 1 1 1 ...
$ SD_Female : Factor w/ 29 levels ""," 1.25",".63",..: 16 20 2 10 11 4 3 1 1 1 ...
$ Mean_Men : Factor w/ 33 levels "","-0.05","/",..: 29 22 10 27 28 25 26 1 1 1 ...
$ SD_Men : Factor w/ 30 levels "",".75",".86",..: 17 20 22 9 8 3 2 1 1 1 ...
$ Mean_Gender : Factor w/ 12 levels "",".20",".27",..: 10 10 10 10 10 10 10 1 1 4 ...
$ SD_Gender : Factor w/ 9 levels "",".40",".43",..: 7 7 7 7 7 7 7 1 1 5 ...
$ Mean_UB : Factor w/ 16 levels "","-.36",".22",..: 4 4 4 4 4 4 4 1 1 12 ...
$ SD_UB : Factor w/ 15 levels "",".42",".95",..: 4 4 4 4 4 4 4 1 1 13 ...
$ Frequency_Use_UB : Factor w/ 7 levels "","/","M: 13/51 (25%)\nF: 5/45 (11%)",..: 2 2 2 2 2 2 2 1 1 2 ...
$ Correlation_Gender_UB : Factor w/ 16 levels "","-.02\n","-.03",..: 15 15 15 15 15 15 15 1 1 10 ...

Ich habe die Liste etwas ausgedünnt. Bei einigen Variablen bin ich mir nicht ganz sicher, ob es Faktoren oder Zahlen sein sollen.
Aber z.B. die MIttelwerte und die Standardabweichungen sollen garantiert keine Faktoren sein.

Es ist immens wichtig, diese Kontrolle nach dem Einlesen der Daten durchzuführen

Gruß, Jörg

Danke für die Hilfe!!
Habe das Problem gelöst, indem ich alle / und ^ etc. aus der Tabelle gelöscht habe. Jetzt funktioniert es.