Multilevel modeling with three levels of data analysis

Question

I am trying to perform a 3-level multilevel model to identify factors that shape practice; Zone Level, HH level, and Individual level.

Here is a snip of my dataset

ZoneID  Zone    HouseID Access  PersonID    Age     Gender  Education   Practice
UZ  Urban   HH1 No  P1  31  Female  Secondary   No
UZ  Urban   HH2 No  P2  33  Female  Tertiary    No
UZ  Urban   HH3 Yes P3  44  Female  Tertiary    Yes
UZ  Urban   HH4 No  P4  28  Male    Tertiary    No
UZ  Urban   HH5 Yes P5  38  Female  Tertiary    No
UZ  Urban   HH6 No  P6  33  Female  Secondary   No
UZ  Urban   HH7 Yes P7  45  Female  Secondary   Yes
UZ  Urban   HH8 Yes P8  20  Female  Secondary   No
UZ  Urban   HH9 Yes P9  44  Female  Tertiary    No
UZ  Urban   HH9 Yes P10 19  Female  Primary No
UZ  Urban   HH10    No  P11 26  Female  Secondary   No
UZ  Urban   HH10    No  P12 26  Female  Secondary   No
UZ  Urban   HH10    No  P13 32  Male    Secondary   No
UZ  Urban   HH11    No  P14 25  Female  Secondary   No
UZ  Urban   HH11    No  P15 36  Male    Secondary   No
UZ  Urban   HH12    Yes P16 47  Female  Secondary   Yes
UZ  Urban   HH13    Yes P17 24  Female  Secondary   Yes
UZ  Urban   HH13    Yes P18 38  Female  Secondary   Yes
UZ  Urban   HH13    Yes P19 23  Female  Secondary   Yes
UZ  Urban   HH13    Yes P20 46  Male    Secondary   No
UZ  Urban   HH13    Yes P21 45  Male    Secondary   No
UZ  Urban   HH13    Yes P22 40  Male    Secondary   No
UZ  Urban   HH14    Yes P23 53  Female  Secondary   No
UZ  Urban   HH15    Yes P24 64  Female  Tertiary    Yes
UZ  Urban   HH15    Yes P25 68  Male    Tertiary    No
UZ  Urban   HH16    Yes P26 55  Female  No education    Yes
UZ  Urban   HH17    Yes P27 28  Female  Secondary   Yes
UZ  Urban   HH18    Yes P28 56  Female  Secondary   No
UZ  Urban   HH19    Yes P29 39  Female  Tertiary    Yes
UZ  Urban   HH19    Yes P30 46  Male    Secondary   No
UZ  Urban   HH19    Yes P31 11  Male    Primary No
UZ  Urban   HH20    Yes P32 57  Female  Secondary   Yes
UZ  Urban   HH20    Yes P33 54  Female  Tertiary    Yes
UZ  Urban   HH20    Yes P34 28  Female  Tertiary    Yes
UZ  Urban   HH20    Yes P35 13  Female  Primary Yes
UZ  Urban   HH20    Yes P36 10  Male    Primary No
UZ  Urban   HH20    Yes P37 23  Male    Primary No
UZ  Urban   HH21    Yes P38 42  Female  Tertiary    No
UZ  Urban   HH21    Yes P39 37  Male    Tertiary    No
UZ  Urban   HH22    Yes P40 43  Male    Secondary   No
PUZ Peri-urban  HH23    Yes P41 78  Female  No education    Yes
PUZ Peri-urban  HH23    Yes P42 36  Female  Secondary   Yes
PUZ Peri-urban  HH23    Yes P43 16  Female  Secondary   Yes
PUZ Peri-urban  HH23    Yes P44 15  Female  Primary Yes
PUZ Peri-urban  HH23    Yes P45 80  Male    Primary No
PUZ Peri-urban  HH23    Yes P46 14  Male    Primary No
PUZ Peri-urban  HH24    Yes P47 76  Female  Tertiary    Yes
PUZ Peri-urban  HH24    Yes P48 34  Female  Tertiary    Yes
PUZ Peri-urban  HH25    Yes P49 73  Female  Tertiary    Yes
PUZ Peri-urban  HH25    Yes P50 29  Female  Secondary   Yes
PUZ Peri-urban  HH25    Yes P51 10  Female  Secondary   Yes
PUZ Peri-urban  HH26    Yes P52 77  Female  Tertiary    Yes
PUZ Peri-urban  HH26    Yes P53 15  Female  Secondary   Yes
PUZ Peri-urban  HH26    Yes P54 13  Female  Primary Yes
PUZ Peri-urban  HH26    Yes P55 13  Female  Primary Yes
PUZ Peri-urban  HH26    Yes P56 10  Female  Primary No
PUZ Peri-urban  HH26    Yes P57 30  Male    Tertiary    No
PUZ Peri-urban  HH26    Yes P58 24  Male    Tertiary    No
PUZ Peri-urban  HH26    Yes P59 44  Male    Secondary   No
PUZ Peri-urban  HH27    Yes P60 50  Female  Secondary   Yes
PUZ Peri-urban  HH27    Yes P61 21  Male    Secondary   No
PUZ Peri-urban  HH28    Yes P62 48  Female  Secondary   Yes
PUZ Peri-urban  HH28    Yes P63 64  Male    Primary No
PUZ Peri-urban  HH29    Yes P64 63  Female  Secondary   Yes
PUZ Peri-urban  HH29    Yes P65 15  Female  Primary Yes
PUZ Peri-urban  HH29    Yes P66 14  Female  Primary Yes
PUZ Peri-urban  HH29    Yes P67 10  Female  Primary Yes
PUZ Peri-urban  HH29    Yes P68 43  Male    Primary Yes
PUZ Peri-urban  HH29    Yes P69 23  Male    Primary Yes
PUZ Peri-urban  HH30    Yes P70 60  Female  Tertiary    Yes
PUZ Peri-urban  HH30    Yes P71 37  Female  No education    Yes
PUZ Peri-urban  HH31    Yes P72 44  Male    Primary Yes
PUZ Peri-urban  HH32    Yes P73 55  Female  Secondary   Yes
PUZ Peri-urban  HH32    Yes P74 52  Female  Tertiary    Yes
PUZ Peri-urban  HH32    Yes P75 26  Female  Secondary   Yes
PUZ Peri-urban  HH32    Yes P76 18  Female  Primary Yes
PUZ Peri-urban  HH32    Yes P77 18  Female  Secondary   Yes
PUZ Peri-urban  HH32    Yes P78 53  Male    Primary Yes
PUZ Peri-urban  HH33    Yes P79 49  Female  Secondary   Yes
PUZ Peri-urban  HH33    Yes P80 50  Male    Secondary   No
PUZ Peri-urban  HH34    Yes P81 72  Female  Primary Yes
PUZ Peri-urban  HH34    Yes P82 66  Female  Primary Yes
PUZ Peri-urban  HH34    Yes P83 39  Female  Secondary   Yes
PUZ Peri-urban  HH34    Yes P84 25  Female  Tertiary    Yes
PUZ Peri-urban  HH34    Yes P85 27  Male    Secondary   No
PUZ Peri-urban  HH34    Yes P86 11  Male    Primary No
PUZ Peri-urban  HH35    Yes P87 29  Female  Tertiary    Yes
PUZ Peri-urban  HH35    Yes P88 59  Male    No education    No
PUZ Peri-urban  HH36    Yes P89 44  Female  Secondary   Yes
PUZ Peri-urban  HH36    Yes P90 13  Female  Primary Yes
PUZ Peri-urban  HH37    Yes P91 88  Female  Primary Yes
PUZ Peri-urban  HH37    Yes P92 45  Female  Secondary   Yes
PUZ Peri-urban  HH37    Yes P93 10  Female  Primary Yes
PUZ Peri-urban  HH37    Yes P94 87  Male    Primary No
PUZ Peri-urban  HH37    Yes P95 11  Male    Primary No
RZ  Rural   HH38    Yes P96 62  Female  Secondary   Yes
RZ  Rural   HH38    Yes P97 28  Female  Tertiary    Yes
RZ  Rural   HH38    Yes P98 21  Male    Secondary   Yes
RZ  Rural   HH38    Yes P99 18  Male    Primary Yes
RZ  Rural   HH38    Yes P100    13  Male    Primary No
RZ  Rural   HH39    Yes P101    65  Female  Secondary   Yes
RZ  Rural   HH39    Yes P102    66  Male    Secondary   No
RZ  Rural   HH39    Yes P103    20  Male    Primary No
RZ  Rural   HH39    Yes P104    16  Male    Secondary   No
RZ  Rural   HH40    No  P105    30  Female  Secondary   No
RZ  Rural   HH40    No  P106    18  Female  Secondary   No
RZ  Rural   HH40    No  P107    38  Male    Secondary   No
RZ  Rural   HH41    Yes P108    96  Female  No education    Yes
RZ  Rural   HH41    Yes P109    85  Female  No education    Yes
RZ  Rural   HH41    Yes P110    57  Female  Tertiary    Yes
RZ  Rural   HH41    Yes P111    28  Female  Secondary   Yes
RZ  Rural   HH41    Yes P112    15  Female  Secondary   Yes
RZ  Rural   HH41    Yes P113    12  Female  Primary Yes
RZ  Rural   HH41    Yes P114    37  Male    Secondary   Yes
RZ  Rural   HH41    Yes P115    32  Male    Secondary   Yes
RZ  Rural   HH41    Yes P116    15  Male    Secondary   Yes
RZ  Rural   HH42    Yes P117    80  Female  No education    Yes
RZ  Rural   HH42    Yes P118    40  Female  Secondary   Yes
RZ  Rural   HH42    Yes P119    39  Male    Secondary   No
RZ  Rural   HH43    Yes P120    33  Female  Secondary   Yes

I have been trying to find a code that I can relate to and build on to address my research question with no success.

This is what I have done but it does not seem to be complete as it appears to only have the individual level, which I am not sure is analyse properly too, and I am unsure of how to reproduce the model for the three levels, to include the HH and Zone level:

model <- glmer(
  Practice ~ Age + Gender + Education + (1 | ZoneID/HouseID),
  data = OshingaliMLM,
  family = binomial(link = "logit")

The results:

    Generalized linear mixed model fit by maximum likelihood (Laplace Approximation) ['glmerMod']
 Family: binomial  ( logit )
Formula: Practice ~ Zone + Access + Marketing + Age + Gender + Education +      Employnment + (1 | ZoneID/HouseID)
   Data: OshingaliMLM

     AIC      BIC   logLik deviance df.resid 
   240.4    304.6   -104.2    208.4      393 

Scaled residuals: 
    Min      1Q  Median      3Q     Max 
-6.9779 -0.0132  0.0112  0.1063  2.6693 

Random effects:
 Groups         Name        Variance  Std.Dev.
 HouseID:ZoneID (Intercept) 1.517e+01 3.8946  
 ZoneID         (Intercept) 3.961e-04 0.0199  
Number of obs: 409, groups:  HouseID:ZoneID, 120; ZoneID, 4

Fixed effects:
                       Estimate Std. Error z value Pr(>|z|)    
(Intercept)            -4.31056   11.01055  -0.391  0.69543    
ZoneRemote rural        2.91666    1.88224   1.550  0.12124    
ZoneRural               0.13047    1.57362   0.083  0.93392    
ZoneUrban              -6.59533    2.24895  -2.933  0.00336 ** 
AccessYes              19.19813   24.22117   0.793  0.42800    
MarketingYes          -12.09891   26.44908  -0.457  0.64735    
Age                     0.01981    0.03214   0.616  0.53760    
GenderMale             -8.89134    1.72704  -5.148 2.63e-07 ***
EducationPrimary        1.62540    1.39860   1.162  0.24517    
EducationSecondary      2.78063    1.51940   1.830  0.06724 .  
EducationTertiary       3.48391    2.05194   1.698  0.08953 .  
EmploynmentLeaners      1.89690    2.28033   0.832  0.40549    
EmploynmentPensioners  -1.09713    2.11383  -0.519  0.60374    
EmploynmentUnemployed   2.49564    1.95557   1.276  0.20189    
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Correlation matrix not shown by default, as p = 14 > 12.
Use print(x, correlation=TRUE)  or
    vcov(x)        if you need it

optimizer (Nelder_Mead) convergence code: 4 (failure to converge in 10000 evaluations)
Model failed to converge with max|grad| = 0.753382 (tol = 0.002, component 1)
Model is nearly unidentifiable: large eigenvalue ratio
 - Rescale variables?
failure to converge in 10000 evaluations